Upload 42 files

.gitattributes

@@ -33,3 +33,6 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+qwen3_8b_sft_rtl/checkpoint-2600/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+qwen3_8b_sft_rtl/checkpoint-2757/tokenizer.json filter=lfs diff=lfs merge=lfs -text
+qwen3_8b_sft_rtl/tokenizer.json filter=lfs diff=lfs merge=lfs -text

qwen3_8b_sft_rtl/README.md

@@ -0,0 +1,81 @@
+---
+library_name: peft
+license: other
+base_model: /workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218
+tags:
+- base_model:adapter:/workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218
+- llama-factory
+- lora
+- transformers
+pipeline_tag: text-generation
+model-index:
+- name: metrex_qwen3_8b_lora_lte1000
+  results: []
+---
+
+<!-- This model card has been generated automatically according to the information the Trainer had access to. You
+should probably proofread and complete it, then remove this comment. -->
+
+# metrex_qwen3_8b_lora_lte1000
+
+This model is a fine-tuned version of [/workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218](https://huggingface.co//workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218) on the metrex_small_lte1000 dataset.
+It achieves the following results on the evaluation set:
+- Loss: 0.0671
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 0.0002
+- train_batch_size: 2
+- eval_batch_size: 2
+- seed: 42
+- distributed_type: multi-GPU
+- num_devices: 4
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 16
+- total_eval_batch_size: 8
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
+- lr_scheduler_type: cosine
+- lr_scheduler_warmup_steps: 100
+- num_epochs: 3.0
+
+### Training results
+
+| Training Loss | Epoch  | Step | Validation Loss |
+|:-------------:|:------:|:----:|:---------------:|
+| 0.0949        | 0.2176 | 200  | 0.0979          |
+| 0.0899        | 0.4353 | 400  | 0.0907          |
+| 0.0786        | 0.6529 | 600  | 0.0875          |
+| 0.0827        | 0.8705 | 800  | 0.0835          |
+| 0.0763        | 1.0881 | 1000 | 0.0822          |
+| 0.0723        | 1.3058 | 1200 | 0.0801          |
+| 0.0724        | 1.5234 | 1400 | 0.0773          |
+| 0.0733        | 1.7410 | 1600 | 0.0752          |
+| 0.0708        | 1.9587 | 1800 | 0.0720          |
+| 0.0535        | 2.1763 | 2000 | 0.0703          |
+| 0.0612        | 2.3939 | 2200 | 0.0699          |
+| 0.0612        | 2.6115 | 2400 | 0.0673          |
+| 0.0522        | 2.8292 | 2600 | 0.0671          |
+
+
+### Framework versions
+
+- PEFT 0.18.1
+- Transformers 5.2.0
+- Pytorch 2.6.0+cu124
+- Datasets 4.0.0
+- Tokenizers 0.22.2

qwen3_8b_sft_rtl/adapter_config.json

@@ -0,0 +1,46 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "/workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 512,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 256,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "down_proj",
+    "up_proj",
+    "gate_proj",
+    "q_proj",
+    "o_proj",
+    "k_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

qwen3_8b_sft_rtl/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1af14129d5c06582bdfe78e4509c38e3850eebc975639aa610959c937020540
+size 2793475840

qwen3_8b_sft_rtl/all_results.json

@@ -0,0 +1,12 @@
+{
+    "epoch": 3.0,
+    "eval_loss": 0.06705856323242188,
+    "eval_runtime": 5.4283,
+    "eval_samples_per_second": 14.738,
+    "eval_steps_per_second": 1.842,
+    "total_flos": 4.480981168874848e+18,
+    "train_loss": 0.07706532943797881,
+    "train_runtime": 10631.5322,
+    "train_samples_per_second": 4.149,
+    "train_steps_per_second": 0.259
+}

qwen3_8b_sft_rtl/chat_template.jinja

@@ -0,0 +1,89 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

qwen3_8b_sft_rtl/checkpoint-2600/README.md

@@ -0,0 +1,208 @@
+---
+base_model: /workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:/workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218
+- llama-factory
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.1

qwen3_8b_sft_rtl/checkpoint-2600/adapter_config.json

@@ -0,0 +1,46 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "/workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 512,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 256,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "down_proj",
+    "up_proj",
+    "gate_proj",
+    "q_proj",
+    "o_proj",
+    "k_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

qwen3_8b_sft_rtl/checkpoint-2600/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:26d27e222a509bbbee96f5b47ac1fe8d9c8fdc91163af029fde145594951e92b
+size 2793475840

qwen3_8b_sft_rtl/checkpoint-2600/chat_template.jinja

@@ -0,0 +1,89 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

qwen3_8b_sft_rtl/checkpoint-2600/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4b68ff8255ff0202668e594ed86d2704e6dc6f47e8a87779b2d04cdcf535785a
+size 5587244546

qwen3_8b_sft_rtl/checkpoint-2600/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:20c3df3d1473394f82a079b1531a84087afee9f941b18680334d8708ce688d9f
+size 15024

qwen3_8b_sft_rtl/checkpoint-2600/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:1f7becaf0c70dc7d589513dc631eae2778261cd1607f4c23f887fb03a24b5c25
+size 15024

qwen3_8b_sft_rtl/checkpoint-2600/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:eae5a4d65ceca428b62bce83ab5de4132d867df0be7e4bbae09dd0d4c939cdf3
+size 15024

qwen3_8b_sft_rtl/checkpoint-2600/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:c4a33ec2d09e1834aa85cadd7ea7d97b1cd169fe32902e6ddf9886c2d24697b5
+size 15024

qwen3_8b_sft_rtl/checkpoint-2600/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d22e345017948d40b9ae5db72fc9b3b62516a6f2f02b5607925994484447ede4
+size 1064

qwen3_8b_sft_rtl/checkpoint-2600/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650

qwen3_8b_sft_rtl/checkpoint-2600/tokenizer_config.json

@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

qwen3_8b_sft_rtl/checkpoint-2600/trainer_state.json

@@ -0,0 +1,1958 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.8291621327529923,
+  "eval_steps": 200,
+  "global_step": 2600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01088139281828074,
+      "grad_norm": 0.2948031425476074,
+      "learning_rate": 1.8e-05,
+      "loss": 0.741086769104004,
+      "step": 10
+    },
+    {
+      "epoch": 0.02176278563656148,
+      "grad_norm": 0.15493904054164886,
+      "learning_rate": 3.8e-05,
+      "loss": 0.3913477897644043,
+      "step": 20
+    },
+    {
+      "epoch": 0.03264417845484222,
+      "grad_norm": 0.136435404419899,
+      "learning_rate": 5.8e-05,
+      "loss": 0.22936077117919923,
+      "step": 30
+    },
+    {
+      "epoch": 0.04352557127312296,
+      "grad_norm": 0.09819520264863968,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 0.17587745189666748,
+      "step": 40
+    },
+    {
+      "epoch": 0.0544069640914037,
+      "grad_norm": 0.10766161978244781,
+      "learning_rate": 9.8e-05,
+      "loss": 0.14949668645858766,
+      "step": 50
+    },
+    {
+      "epoch": 0.06528835690968444,
+      "grad_norm": 0.08924636989831924,
+      "learning_rate": 0.000118,
+      "loss": 0.13353660106658935,
+      "step": 60
+    },
+    {
+      "epoch": 0.07616974972796518,
+      "grad_norm": 0.08617673069238663,
+      "learning_rate": 0.000138,
+      "loss": 0.120628023147583,
+      "step": 70
+    },
+    {
+      "epoch": 0.08705114254624592,
+      "grad_norm": 0.06150153651833534,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 0.11396801471710205,
+      "step": 80
+    },
+    {
+      "epoch": 0.09793253536452666,
+      "grad_norm": 0.06216437742114067,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 0.1074080228805542,
+      "step": 90
+    },
+    {
+      "epoch": 0.1088139281828074,
+      "grad_norm": 0.07203802466392517,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 0.10922808647155761,
+      "step": 100
+    },
+    {
+      "epoch": 0.11969532100108814,
+      "grad_norm": 0.04227524250745773,
+      "learning_rate": 0.00019999433803003197,
+      "loss": 0.10778391361236572,
+      "step": 110
+    },
+    {
+      "epoch": 0.1305767138193689,
+      "grad_norm": 0.06469517946243286,
+      "learning_rate": 0.00019997476661131597,
+      "loss": 0.10001739263534545,
+      "step": 120
+    },
+    {
+      "epoch": 0.14145810663764963,
+      "grad_norm": 0.04663902893662453,
+      "learning_rate": 0.0001999412185784487,
+      "loss": 0.10446863174438477,
+      "step": 130
+    },
+    {
+      "epoch": 0.15233949945593037,
+      "grad_norm": 0.03584141284227371,
+      "learning_rate": 0.0001998936986214928,
+      "loss": 0.10438344478607178,
+      "step": 140
+    },
+    {
+      "epoch": 0.1632208922742111,
+      "grad_norm": 0.03158590570092201,
+      "learning_rate": 0.00019983221338380563,
+      "loss": 0.09951521158218384,
+      "step": 150
+    },
+    {
+      "epoch": 0.17410228509249184,
+      "grad_norm": 0.02909119985997677,
+      "learning_rate": 0.0001997567714611106,
+      "loss": 0.09911214709281921,
+      "step": 160
+    },
+    {
+      "epoch": 0.18498367791077258,
+      "grad_norm": 0.030274737626314163,
+      "learning_rate": 0.0001996673834002955,
+      "loss": 0.09334277510643005,
+      "step": 170
+    },
+    {
+      "epoch": 0.19586507072905332,
+      "grad_norm": 0.028922295197844505,
+      "learning_rate": 0.00019956406169793805,
+      "loss": 0.09431872367858887,
+      "step": 180
+    },
+    {
+      "epoch": 0.20674646354733406,
+      "grad_norm": 0.030195271596312523,
+      "learning_rate": 0.00019944682079855867,
+      "loss": 0.10089273452758789,
+      "step": 190
+    },
+    {
+      "epoch": 0.2176278563656148,
+      "grad_norm": 0.029216868802905083,
+      "learning_rate": 0.0001993156770926014,
+      "loss": 0.0949325442314148,
+      "step": 200
+    },
+    {
+      "epoch": 0.2176278563656148,
+      "eval_loss": 0.09785260260105133,
+      "eval_runtime": 5.4576,
+      "eval_samples_per_second": 14.659,
+      "eval_steps_per_second": 1.832,
+      "step": 200
+    },
+    {
+      "epoch": 0.22850924918389554,
+      "grad_norm": 0.029826847836375237,
+      "learning_rate": 0.00019917064891414232,
+      "loss": 0.09331356287002564,
+      "step": 210
+    },
+    {
+      "epoch": 0.23939064200217627,
+      "grad_norm": 0.03143204376101494,
+      "learning_rate": 0.00019901175653832645,
+      "loss": 0.09654195308685302,
+      "step": 220
+    },
+    {
+      "epoch": 0.250272034820457,
+      "grad_norm": 0.02917400933802128,
+      "learning_rate": 0.00019883902217853326,
+      "loss": 0.09114633202552795,
+      "step": 230
+    },
+    {
+      "epoch": 0.2611534276387378,
+      "grad_norm": 0.033463530242443085,
+      "learning_rate": 0.0001986524699832712,
+      "loss": 0.09384369254112243,
+      "step": 240
+    },
+    {
+      "epoch": 0.2720348204570185,
+      "grad_norm": 0.027332467958331108,
+      "learning_rate": 0.00019845212603280174,
+      "loss": 0.08879860639572143,
+      "step": 250
+    },
+    {
+      "epoch": 0.28291621327529926,
+      "grad_norm": 0.03050387091934681,
+      "learning_rate": 0.00019823801833549327,
+      "loss": 0.09065196514129639,
+      "step": 260
+    },
+    {
+      "epoch": 0.29379760609357997,
+      "grad_norm": 0.024866925552487373,
+      "learning_rate": 0.00019801017682390552,
+      "loss": 0.09161882400512696,
+      "step": 270
+    },
+    {
+      "epoch": 0.30467899891186073,
+      "grad_norm": 0.037345923483371735,
+      "learning_rate": 0.0001977686333506049,
+      "loss": 0.09159514904022217,
+      "step": 280
+    },
+    {
+      "epoch": 0.31556039173014144,
+      "grad_norm": 0.0632031187415123,
+      "learning_rate": 0.00019751342168371154,
+      "loss": 0.0902558445930481,
+      "step": 290
+    },
+    {
+      "epoch": 0.3264417845484222,
+      "grad_norm": 0.029686614871025085,
+      "learning_rate": 0.00019724457750217839,
+      "loss": 0.09181964993476868,
+      "step": 300
+    },
+    {
+      "epoch": 0.3373231773667029,
+      "grad_norm": 0.03043133206665516,
+      "learning_rate": 0.0001969621383908033,
+      "loss": 0.08723740577697754,
+      "step": 310
+    },
+    {
+      "epoch": 0.3482045701849837,
+      "grad_norm": 0.041717205196619034,
+      "learning_rate": 0.00019666614383497457,
+      "loss": 0.09164115190505981,
+      "step": 320
+    },
+    {
+      "epoch": 0.3590859630032644,
+      "grad_norm": 0.03025348111987114,
+      "learning_rate": 0.00019635663521515087,
+      "loss": 0.08903414011001587,
+      "step": 330
+    },
+    {
+      "epoch": 0.36996735582154516,
+      "grad_norm": 0.030753403902053833,
+      "learning_rate": 0.00019603365580107615,
+      "loss": 0.08611454963684081,
+      "step": 340
+    },
+    {
+      "epoch": 0.3808487486398259,
+      "grad_norm": 0.021834490820765495,
+      "learning_rate": 0.0001956972507457306,
+      "loss": 0.08958455920219421,
+      "step": 350
+    },
+    {
+      "epoch": 0.39173014145810664,
+      "grad_norm": 0.025724470615386963,
+      "learning_rate": 0.000195347467079018,
+      "loss": 0.08454820513725281,
+      "step": 360
+    },
+    {
+      "epoch": 0.40261153427638735,
+      "grad_norm": 0.034213438630104065,
+      "learning_rate": 0.00019498435370119094,
+      "loss": 0.09067847728729247,
+      "step": 370
+    },
+    {
+      "epoch": 0.4134929270946681,
+      "grad_norm": 0.0437682569026947,
+      "learning_rate": 0.00019460796137601467,
+      "loss": 0.0873767077922821,
+      "step": 380
+    },
+    {
+      "epoch": 0.4243743199129489,
+      "grad_norm": 0.027341296896338463,
+      "learning_rate": 0.00019421834272367,
+      "loss": 0.08544692993164063,
+      "step": 390
+    },
+    {
+      "epoch": 0.4352557127312296,
+      "grad_norm": 0.030272383242845535,
+      "learning_rate": 0.00019381555221339707,
+      "loss": 0.08988077044487,
+      "step": 400
+    },
+    {
+      "epoch": 0.4352557127312296,
+      "eval_loss": 0.09069998562335968,
+      "eval_runtime": 5.3759,
+      "eval_samples_per_second": 14.881,
+      "eval_steps_per_second": 1.86,
+      "step": 400
+    },
+    {
+      "epoch": 0.44613710554951036,
+      "grad_norm": 0.030490044504404068,
+      "learning_rate": 0.00019339964615588043,
+      "loss": 0.088669753074646,
+      "step": 410
+    },
+    {
+      "epoch": 0.45701849836779107,
+      "grad_norm": 0.030257854610681534,
+      "learning_rate": 0.0001929706826953767,
+      "loss": 0.09294685125350952,
+      "step": 420
+    },
+    {
+      "epoch": 0.46789989118607184,
+      "grad_norm": 0.03422554209828377,
+      "learning_rate": 0.00019252872180158605,
+      "loss": 0.08597595095634461,
+      "step": 430
+    },
+    {
+      "epoch": 0.47878128400435255,
+      "grad_norm": 0.026077628135681152,
+      "learning_rate": 0.0001920738252612682,
+      "loss": 0.08706499338150024,
+      "step": 440
+    },
+    {
+      "epoch": 0.4896626768226333,
+      "grad_norm": 0.025868231430649757,
+      "learning_rate": 0.0001916060566696045,
+      "loss": 0.08425920009613037,
+      "step": 450
+    },
+    {
+      "epoch": 0.500544069640914,
+      "grad_norm": 0.03231549635529518,
+      "learning_rate": 0.00019112548142130757,
+      "loss": 0.08545957803726197,
+      "step": 460
+    },
+    {
+      "epoch": 0.5114254624591947,
+      "grad_norm": 0.03557935357093811,
+      "learning_rate": 0.00019063216670147854,
+      "loss": 0.08239468932151794,
+      "step": 470
+    },
+    {
+      "epoch": 0.5223068552774756,
+      "grad_norm": 0.023028500378131866,
+      "learning_rate": 0.00019012618147621488,
+      "loss": 0.08288769721984864,
+      "step": 480
+    },
+    {
+      "epoch": 0.5331882480957563,
+      "grad_norm": 0.021350380033254623,
+      "learning_rate": 0.00018960759648296865,
+      "loss": 0.08916928768157958,
+      "step": 490
+    },
+    {
+      "epoch": 0.544069640914037,
+      "grad_norm": 0.028527546674013138,
+      "learning_rate": 0.0001890764842206573,
+      "loss": 0.09119058847427368,
+      "step": 500
+    },
+    {
+      "epoch": 0.5549510337323177,
+      "grad_norm": 0.023469405248761177,
+      "learning_rate": 0.00018853291893952833,
+      "loss": 0.08246206045150757,
+      "step": 510
+    },
+    {
+      "epoch": 0.5658324265505985,
+      "grad_norm": 0.028494561091065407,
+      "learning_rate": 0.0001879769766307789,
+      "loss": 0.0803530752658844,
+      "step": 520
+    },
+    {
+      "epoch": 0.5767138193688792,
+      "grad_norm": 0.027940815314650536,
+      "learning_rate": 0.00018740873501593214,
+      "loss": 0.08848085403442382,
+      "step": 530
+    },
+    {
+      "epoch": 0.5875952121871599,
+      "grad_norm": 0.02026485837996006,
+      "learning_rate": 0.0001868282735359717,
+      "loss": 0.08710720539093017,
+      "step": 540
+    },
+    {
+      "epoch": 0.5984766050054406,
+      "grad_norm": 0.026024585589766502,
+      "learning_rate": 0.00018623567334023563,
+      "loss": 0.07982800006866456,
+      "step": 550
+    },
+    {
+      "epoch": 0.6093579978237215,
+      "grad_norm": 0.029224518686532974,
+      "learning_rate": 0.0001856310172750717,
+      "loss": 0.08492777943611145,
+      "step": 560
+    },
+    {
+      "epoch": 0.6202393906420022,
+      "grad_norm": 0.02598229981958866,
+      "learning_rate": 0.00018501438987225536,
+      "loss": 0.08290873765945435,
+      "step": 570
+    },
+    {
+      "epoch": 0.6311207834602829,
+      "grad_norm": 0.023201750591397285,
+      "learning_rate": 0.00018438587733717187,
+      "loss": 0.08711419701576233,
+      "step": 580
+    },
+    {
+      "epoch": 0.6420021762785637,
+      "grad_norm": 0.026234455406665802,
+      "learning_rate": 0.00018374556753676506,
+      "loss": 0.08342958688735962,
+      "step": 590
+    },
+    {
+      "epoch": 0.6528835690968444,
+      "grad_norm": 0.021891364827752113,
+      "learning_rate": 0.00018309354998725306,
+      "loss": 0.07860854864120484,
+      "step": 600
+    },
+    {
+      "epoch": 0.6528835690968444,
+      "eval_loss": 0.08754301071166992,
+      "eval_runtime": 5.5341,
+      "eval_samples_per_second": 14.456,
+      "eval_steps_per_second": 1.807,
+      "step": 600
+    },
+    {
+      "epoch": 0.6637649619151251,
+      "grad_norm": 0.026508808135986328,
+      "learning_rate": 0.00018242991584161393,
+      "loss": 0.08734351992607117,
+      "step": 610
+    },
+    {
+      "epoch": 0.6746463547334058,
+      "grad_norm": 0.02286313846707344,
+      "learning_rate": 0.0001817547578768424,
+      "loss": 0.08437882661819458,
+      "step": 620
+    },
+    {
+      "epoch": 0.6855277475516867,
+      "grad_norm": 0.02370450273156166,
+      "learning_rate": 0.00018106817048097937,
+      "loss": 0.08016583323478699,
+      "step": 630
+    },
+    {
+      "epoch": 0.6964091403699674,
+      "grad_norm": 0.025123964995145798,
+      "learning_rate": 0.0001803702496399165,
+      "loss": 0.076961088180542,
+      "step": 640
+    },
+    {
+      "epoch": 0.7072905331882481,
+      "grad_norm": 0.019611909985542297,
+      "learning_rate": 0.0001796610929239771,
+      "loss": 0.08545655012130737,
+      "step": 650
+    },
+    {
+      "epoch": 0.7181719260065288,
+      "grad_norm": 0.020668957382440567,
+      "learning_rate": 0.00017894079947427573,
+      "loss": 0.0809054434299469,
+      "step": 660
+    },
+    {
+      "epoch": 0.7290533188248096,
+      "grad_norm": 0.028357358649373055,
+      "learning_rate": 0.0001782094699888582,
+      "loss": 0.07935836911201477,
+      "step": 670
+    },
+    {
+      "epoch": 0.7399347116430903,
+      "grad_norm": 0.018039165064692497,
+      "learning_rate": 0.00017746720670862363,
+      "loss": 0.08474135994911194,
+      "step": 680
+    },
+    {
+      "epoch": 0.750816104461371,
+      "grad_norm": 0.02181115373969078,
+      "learning_rate": 0.0001767141134030313,
+      "loss": 0.08088613748550415,
+      "step": 690
+    },
+    {
+      "epoch": 0.7616974972796517,
+      "grad_norm": 0.029043374583125114,
+      "learning_rate": 0.00017595029535559336,
+      "loss": 0.078003191947937,
+      "step": 700
+    },
+    {
+      "epoch": 0.7725788900979326,
+      "grad_norm": 0.024250885471701622,
+      "learning_rate": 0.00017517585934915595,
+      "loss": 0.08291269540786743,
+      "step": 710
+    },
+    {
+      "epoch": 0.7834602829162133,
+      "grad_norm": 0.028013000264763832,
+      "learning_rate": 0.00017439091365097106,
+      "loss": 0.08375194072723388,
+      "step": 720
+    },
+    {
+      "epoch": 0.794341675734494,
+      "grad_norm": 0.030651913955807686,
+      "learning_rate": 0.00017359556799756052,
+      "loss": 0.08069210052490235,
+      "step": 730
+    },
+    {
+      "epoch": 0.8052230685527747,
+      "grad_norm": 0.028212785720825195,
+      "learning_rate": 0.00017278993357937455,
+      "loss": 0.08342626094818115,
+      "step": 740
+    },
+    {
+      "epoch": 0.8161044613710555,
+      "grad_norm": 0.024551352486014366,
+      "learning_rate": 0.0001719741230252473,
+      "loss": 0.0814591646194458,
+      "step": 750
+    },
+    {
+      "epoch": 0.8269858541893362,
+      "grad_norm": 0.022670986130833626,
+      "learning_rate": 0.00017114825038665123,
+      "loss": 0.08308478593826293,
+      "step": 760
+    },
+    {
+      "epoch": 0.8378672470076169,
+      "grad_norm": 0.03288818150758743,
+      "learning_rate": 0.0001703124311217524,
+      "loss": 0.07733856439590454,
+      "step": 770
+    },
+    {
+      "epoch": 0.8487486398258978,
+      "grad_norm": 0.02053298056125641,
+      "learning_rate": 0.00016946678207926938,
+      "loss": 0.0828587532043457,
+      "step": 780
+    },
+    {
+      "epoch": 0.8596300326441785,
+      "grad_norm": 0.024455999955534935,
+      "learning_rate": 0.0001686114214821375,
+      "loss": 0.07782068252563476,
+      "step": 790
+    },
+    {
+      "epoch": 0.8705114254624592,
+      "grad_norm": 0.022575655952095985,
+      "learning_rate": 0.00016774646891098134,
+      "loss": 0.08273504972457886,
+      "step": 800
+    },
+    {
+      "epoch": 0.8705114254624592,
+      "eval_loss": 0.08349094539880753,
+      "eval_runtime": 5.5064,
+      "eval_samples_per_second": 14.529,
+      "eval_steps_per_second": 1.816,
+      "step": 800
+    },
+    {
+      "epoch": 0.8813928182807399,
+      "grad_norm": 0.02310466580092907,
+      "learning_rate": 0.00016687204528739694,
+      "loss": 0.08457955718040466,
+      "step": 810
+    },
+    {
+      "epoch": 0.8922742110990207,
+      "grad_norm": 0.02504105120897293,
+      "learning_rate": 0.00016598827285704696,
+      "loss": 0.08319224119186401,
+      "step": 820
+    },
+    {
+      "epoch": 0.9031556039173014,
+      "grad_norm": 0.02492046356201172,
+      "learning_rate": 0.00016509527517257042,
+      "loss": 0.08071508407592773,
+      "step": 830
+    },
+    {
+      "epoch": 0.9140369967355821,
+      "grad_norm": 0.023182213306427002,
+      "learning_rate": 0.0001641931770763101,
+      "loss": 0.08242483139038086,
+      "step": 840
+    },
+    {
+      "epoch": 0.9249183895538629,
+      "grad_norm": 0.022171733900904655,
+      "learning_rate": 0.00016328210468285911,
+      "loss": 0.08000438213348389,
+      "step": 850
+    },
+    {
+      "epoch": 0.9357997823721437,
+      "grad_norm": 0.023160789161920547,
+      "learning_rate": 0.00016236218536143013,
+      "loss": 0.08225069642066955,
+      "step": 860
+    },
+    {
+      "epoch": 0.9466811751904244,
+      "grad_norm": 0.021988507360219955,
+      "learning_rate": 0.0001614335477180488,
+      "loss": 0.07968268394470215,
+      "step": 870
+    },
+    {
+      "epoch": 0.9575625680087051,
+      "grad_norm": 0.0209442637860775,
+      "learning_rate": 0.00016049632157757473,
+      "loss": 0.08057799339294433,
+      "step": 880
+    },
+    {
+      "epoch": 0.9684439608269858,
+      "grad_norm": 0.02684466354548931,
+      "learning_rate": 0.00015955063796555137,
+      "loss": 0.07912254929542542,
+      "step": 890
+    },
+    {
+      "epoch": 0.9793253536452666,
+      "grad_norm": 0.022005856037139893,
+      "learning_rate": 0.00015859662908988885,
+      "loss": 0.07660449743270874,
+      "step": 900
+    },
+    {
+      "epoch": 0.9902067464635473,
+      "grad_norm": 0.022977208718657494,
+      "learning_rate": 0.0001576344283223809,
+      "loss": 0.07685725092887878,
+      "step": 910
+    },
+    {
+      "epoch": 1.001088139281828,
+      "grad_norm": 0.020924242213368416,
+      "learning_rate": 0.00015666417018005938,
+      "loss": 0.07933401465415954,
+      "step": 920
+    },
+    {
+      "epoch": 1.0119695321001088,
+      "grad_norm": 0.02789444290101528,
+      "learning_rate": 0.00015568599030638854,
+      "loss": 0.07339736819267273,
+      "step": 930
+    },
+    {
+      "epoch": 1.0228509249183895,
+      "grad_norm": 0.02638629637658596,
+      "learning_rate": 0.00015470002545230193,
+      "loss": 0.07634537220001221,
+      "step": 940
+    },
+    {
+      "epoch": 1.0337323177366704,
+      "grad_norm": 0.029259871691465378,
+      "learning_rate": 0.0001537064134570844,
+      "loss": 0.08004393577575683,
+      "step": 950
+    },
+    {
+      "epoch": 1.0446137105549511,
+      "grad_norm": 0.03044624626636505,
+      "learning_rate": 0.0001527052932291019,
+      "loss": 0.07677503824234008,
+      "step": 960
+    },
+    {
+      "epoch": 1.0554951033732318,
+      "grad_norm": 0.02535896934568882,
+      "learning_rate": 0.00015169680472638218,
+      "loss": 0.07567052841186524,
+      "step": 970
+    },
+    {
+      "epoch": 1.0663764961915125,
+      "grad_norm": 0.02594912238419056,
+      "learning_rate": 0.00015068108893704797,
+      "loss": 0.0768008291721344,
+      "step": 980
+    },
+    {
+      "epoch": 1.0772578890097932,
+      "grad_norm": 0.026760386303067207,
+      "learning_rate": 0.0001496582878596071,
+      "loss": 0.07604062557220459,
+      "step": 990
+    },
+    {
+      "epoch": 1.088139281828074,
+      "grad_norm": 0.02427523396909237,
+      "learning_rate": 0.00014862854448310074,
+      "loss": 0.07631718516349792,
+      "step": 1000
+    },
+    {
+      "epoch": 1.088139281828074,
+      "eval_loss": 0.08224387466907501,
+      "eval_runtime": 5.3767,
+      "eval_samples_per_second": 14.879,
+      "eval_steps_per_second": 1.86,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0990206746463547,
+      "grad_norm": 0.023537471890449524,
+      "learning_rate": 0.00014759200276711316,
+      "loss": 0.0752741813659668,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1099020674646354,
+      "grad_norm": 0.028346385806798935,
+      "learning_rate": 0.00014654880762164624,
+      "loss": 0.07320383191108704,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1207834602829163,
+      "grad_norm": 0.024628447368741035,
+      "learning_rate": 0.00014549910488686078,
+      "loss": 0.08079155683517455,
+      "step": 1030
+    },
+    {
+      "epoch": 1.131664853101197,
+      "grad_norm": 0.024834809824824333,
+      "learning_rate": 0.0001444430413126877,
+      "loss": 0.07409957647323609,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1425462459194777,
+      "grad_norm": 0.03418092057108879,
+      "learning_rate": 0.0001433807645383123,
+      "loss": 0.07469584345817566,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1534276387377584,
+      "grad_norm": 0.022386226803064346,
+      "learning_rate": 0.00014231242307153417,
+      "loss": 0.07519133687019348,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1643090315560392,
+      "grad_norm": 0.023596055805683136,
+      "learning_rate": 0.00014123816626800535,
+      "loss": 0.07806066274642945,
+      "step": 1070
+    },
+    {
+      "epoch": 1.1751904243743199,
+      "grad_norm": 0.025627054274082184,
+      "learning_rate": 0.00014015814431035046,
+      "loss": 0.07402833104133606,
+      "step": 1080
+    },
+    {
+      "epoch": 1.1860718171926006,
+      "grad_norm": 0.021048523485660553,
+      "learning_rate": 0.00013907250818717082,
+      "loss": 0.07758893370628357,
+      "step": 1090
+    },
+    {
+      "epoch": 1.1969532100108813,
+      "grad_norm": 0.028320247307419777,
+      "learning_rate": 0.00013798140967193607,
+      "loss": 0.0765085220336914,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2078346028291622,
+      "grad_norm": 0.027219558134675026,
+      "learning_rate": 0.0001368850013017659,
+      "loss": 0.07446168661117554,
+      "step": 1110
+    },
+    {
+      "epoch": 1.218715995647443,
+      "grad_norm": 0.022490857169032097,
+      "learning_rate": 0.0001357834363561054,
+      "loss": 0.07697376012802123,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2295973884657236,
+      "grad_norm": 0.02330136112868786,
+      "learning_rate": 0.00013467686883529616,
+      "loss": 0.07688432335853576,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2404787812840044,
+      "grad_norm": 0.025878841057419777,
+      "learning_rate": 0.0001335654534390468,
+      "loss": 0.07023751735687256,
+      "step": 1140
+    },
+    {
+      "epoch": 1.251360174102285,
+      "grad_norm": 0.023564936593174934,
+      "learning_rate": 0.0001324493455448059,
+      "loss": 0.0761339783668518,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2622415669205658,
+      "grad_norm": 0.01974901743233204,
+      "learning_rate": 0.0001313287011860399,
+      "loss": 0.07061100006103516,
+      "step": 1160
+    },
+    {
+      "epoch": 1.2731229597388465,
+      "grad_norm": 0.028187213465571404,
+      "learning_rate": 0.00013020367703041934,
+      "loss": 0.07600026726722717,
+      "step": 1170
+    },
+    {
+      "epoch": 1.2840043525571274,
+      "grad_norm": 0.027759388089179993,
+      "learning_rate": 0.0001290744303579167,
+      "loss": 0.07545133829116821,
+      "step": 1180
+    },
+    {
+      "epoch": 1.2948857453754081,
+      "grad_norm": 0.03569910302758217,
+      "learning_rate": 0.00012794111903881818,
+      "loss": 0.07300611138343811,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3057671381936888,
+      "grad_norm": 0.026020081713795662,
+      "learning_rate": 0.00012680390151165337,
+      "loss": 0.07233045101165772,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3057671381936888,
+      "eval_loss": 0.08009672164916992,
+      "eval_runtime": 5.448,
+      "eval_samples_per_second": 14.684,
+      "eval_steps_per_second": 1.836,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3166485310119695,
+      "grad_norm": 0.02722887694835663,
+      "learning_rate": 0.0001256629367610454,
+      "loss": 0.0722274899482727,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3275299238302503,
+      "grad_norm": 0.021629413589835167,
+      "learning_rate": 0.0001245183842954844,
+      "loss": 0.07423741221427918,
+      "step": 1220
+    },
+    {
+      "epoch": 1.338411316648531,
+      "grad_norm": 0.030425263568758965,
+      "learning_rate": 0.00012337040412502823,
+      "loss": 0.07771967649459839,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3492927094668117,
+      "grad_norm": 0.022577406838536263,
+      "learning_rate": 0.00012221915673893267,
+      "loss": 0.07685770392417908,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3601741022850926,
+      "grad_norm": 0.026748720556497574,
+      "learning_rate": 0.00012106480308321498,
+      "loss": 0.0744680941104889,
+      "step": 1250
+    },
+    {
+      "epoch": 1.371055495103373,
+      "grad_norm": 0.024764206260442734,
+      "learning_rate": 0.00011990750453815324,
+      "loss": 0.07823204398155212,
+      "step": 1260
+    },
+    {
+      "epoch": 1.381936887921654,
+      "grad_norm": 0.023767156526446342,
+      "learning_rate": 0.00011874742289572531,
+      "loss": 0.06924179792404175,
+      "step": 1270
+    },
+    {
+      "epoch": 1.3928182807399347,
+      "grad_norm": 0.025964556261897087,
+      "learning_rate": 0.00011758472033698993,
+      "loss": 0.06726616621017456,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4036996735582155,
+      "grad_norm": 0.024118445813655853,
+      "learning_rate": 0.00011641955940941374,
+      "loss": 0.07383161187171935,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4145810663764962,
+      "grad_norm": 0.0229018721729517,
+      "learning_rate": 0.00011525210300414673,
+      "loss": 0.07341222763061524,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4254624591947769,
+      "grad_norm": 0.02409880980849266,
+      "learning_rate": 0.00011408251433325006,
+      "loss": 0.07354844808578491,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4363438520130578,
+      "grad_norm": 0.039964865893125534,
+      "learning_rate": 0.00011291095690687847,
+      "loss": 0.07191305160522461,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4472252448313383,
+      "grad_norm": 0.02640443667769432,
+      "learning_rate": 0.00011173759451042168,
+      "loss": 0.07049686908721924,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4581066376496192,
+      "grad_norm": 0.022137323394417763,
+      "learning_rate": 0.0001105625911816067,
+      "loss": 0.07753136157989501,
+      "step": 1340
+    },
+    {
+      "epoch": 1.4689880304679,
+      "grad_norm": 0.02352975122630596,
+      "learning_rate": 0.0001093861111875652,
+      "loss": 0.07463480234146118,
+      "step": 1350
+    },
+    {
+      "epoch": 1.4798694232861807,
+      "grad_norm": 0.02214525081217289,
+      "learning_rate": 0.0001082083190018688,
+      "loss": 0.06845179796218873,
+      "step": 1360
+    },
+    {
+      "epoch": 1.4907508161044614,
+      "grad_norm": 0.025555888190865517,
+      "learning_rate": 0.00010702937928153533,
+      "loss": 0.07124170064926147,
+      "step": 1370
+    },
+    {
+      "epoch": 1.501632208922742,
+      "grad_norm": 0.02631489560008049,
+      "learning_rate": 0.00010584945684400963,
+      "loss": 0.07157056331634522,
+      "step": 1380
+    },
+    {
+      "epoch": 1.512513601741023,
+      "grad_norm": 0.026430578902363777,
+      "learning_rate": 0.0001046687166441217,
+      "loss": 0.07220081686973571,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5233949945593035,
+      "grad_norm": 0.024772843345999718,
+      "learning_rate": 0.00010348732375102601,
+      "loss": 0.07243863940238952,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5233949945593035,
+      "eval_loss": 0.07733941823244095,
+      "eval_runtime": 5.3743,
+      "eval_samples_per_second": 14.886,
+      "eval_steps_per_second": 1.861,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5342763873775844,
+      "grad_norm": 0.0237529706209898,
+      "learning_rate": 0.00010230544332512442,
+      "loss": 0.06972206830978393,
+      "step": 1410
+    },
+    {
+      "epoch": 1.545157780195865,
+      "grad_norm": 0.02250276878476143,
+      "learning_rate": 0.00010112324059497649,
+      "loss": 0.07575106620788574,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5560391730141458,
+      "grad_norm": 0.02346464991569519,
+      "learning_rate": 9.994088083420044e-05,
+      "loss": 0.07188562154769898,
+      "step": 1430
+    },
+    {
+      "epoch": 1.5669205658324266,
+      "grad_norm": 0.023935239762067795,
+      "learning_rate": 9.875852933836765e-05,
+      "loss": 0.0751798152923584,
+      "step": 1440
+    },
+    {
+      "epoch": 1.5778019586507073,
+      "grad_norm": 0.02204659767448902,
+      "learning_rate": 9.757635140189394e-05,
+      "loss": 0.06870698928833008,
+      "step": 1450
+    },
+    {
+      "epoch": 1.588683351468988,
+      "grad_norm": 0.022586360573768616,
+      "learning_rate": 9.639451229493133e-05,
+      "loss": 0.07310816049575805,
+      "step": 1460
+    },
+    {
+      "epoch": 1.5995647442872687,
+      "grad_norm": 0.02241821214556694,
+      "learning_rate": 9.521317724026299e-05,
+      "loss": 0.07275789976119995,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6104461371055496,
+      "grad_norm": 0.019981661811470985,
+      "learning_rate": 9.403251139020486e-05,
+      "loss": 0.07513962984085083,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6213275299238301,
+      "grad_norm": 0.0203393641859293,
+      "learning_rate": 9.285267980351719e-05,
+      "loss": 0.0734050989151001,
+      "step": 1490
+    },
+    {
+      "epoch": 1.632208922742111,
+      "grad_norm": 0.027680471539497375,
+      "learning_rate": 9.167384742232898e-05,
+      "loss": 0.06607317924499512,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6430903155603918,
+      "grad_norm": 0.026372363790869713,
+      "learning_rate": 9.049617904907877e-05,
+      "loss": 0.07146948575973511,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6539717083786725,
+      "grad_norm": 0.02281905524432659,
+      "learning_rate": 8.931983932347533e-05,
+      "loss": 0.07334067225456238,
+      "step": 1520
+    },
+    {
+      "epoch": 1.6648531011969532,
+      "grad_norm": 0.02359774336218834,
+      "learning_rate": 8.814499269948053e-05,
+      "loss": 0.0697287678718567,
+      "step": 1530
+    },
+    {
+      "epoch": 1.6757344940152339,
+      "grad_norm": 0.02269565500319004,
+      "learning_rate": 8.697180342231849e-05,
+      "loss": 0.07186954021453858,
+      "step": 1540
+    },
+    {
+      "epoch": 1.6866158868335148,
+      "grad_norm": 0.028245631605386734,
+      "learning_rate": 8.580043550551393e-05,
+      "loss": 0.06877661943435669,
+      "step": 1550
+    },
+    {
+      "epoch": 1.6974972796517953,
+      "grad_norm": 0.02352772280573845,
+      "learning_rate": 8.463105270796276e-05,
+      "loss": 0.07183843851089478,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7083786724700762,
+      "grad_norm": 0.023996759206056595,
+      "learning_rate": 8.346381851103842e-05,
+      "loss": 0.07082110047340393,
+      "step": 1570
+    },
+    {
+      "epoch": 1.719260065288357,
+      "grad_norm": 0.025422753766179085,
+      "learning_rate": 8.229889609573699e-05,
+      "loss": 0.06895624399185181,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7301414581066377,
+      "grad_norm": 0.02499948814511299,
+      "learning_rate": 8.113644831986416e-05,
+      "loss": 0.0707672655582428,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7410228509249184,
+      "grad_norm": 0.030967354774475098,
+      "learning_rate": 7.997663769526753e-05,
+      "loss": 0.07331972122192383,
+      "step": 1600
+    },
+    {
+      "epoch": 1.7410228509249184,
+      "eval_loss": 0.07515527307987213,
+      "eval_runtime": 5.3812,
+      "eval_samples_per_second": 14.867,
+      "eval_steps_per_second": 1.858,
+      "step": 1600
+    },
+    {
+      "epoch": 1.751904243743199,
+      "grad_norm": 0.026022454723715782,
+      "learning_rate": 7.881962636511715e-05,
+      "loss": 0.07261155843734741,
+      "step": 1610
+    },
+    {
+      "epoch": 1.76278563656148,
+      "grad_norm": 0.03104907087981701,
+      "learning_rate": 7.766557608123783e-05,
+      "loss": 0.07076815962791443,
+      "step": 1620
+    },
+    {
+      "epoch": 1.7736670293797605,
+      "grad_norm": 0.026025639846920967,
+      "learning_rate": 7.651464818149587e-05,
+      "loss": 0.0643875777721405,
+      "step": 1630
+    },
+    {
+      "epoch": 1.7845484221980414,
+      "grad_norm": 0.022654540836811066,
+      "learning_rate": 7.536700356724379e-05,
+      "loss": 0.07072726488113404,
+      "step": 1640
+    },
+    {
+      "epoch": 1.795429815016322,
+      "grad_norm": 0.02240171656012535,
+      "learning_rate": 7.422280268082629e-05,
+      "loss": 0.07485873699188232,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8063112078346029,
+      "grad_norm": 0.02509446069598198,
+      "learning_rate": 7.308220548314981e-05,
+      "loss": 0.06636062860488892,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8171926006528836,
+      "grad_norm": 0.023968016728758812,
+      "learning_rate": 7.194537143132023e-05,
+      "loss": 0.06702766418457032,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8280739934711643,
+      "grad_norm": 0.027016611769795418,
+      "learning_rate": 7.081245945635013e-05,
+      "loss": 0.07087734937667847,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8389553862894452,
+      "grad_norm": 0.02593705803155899,
+      "learning_rate": 6.968362794094027e-05,
+      "loss": 0.06891222596168518,
+      "step": 1690
+    },
+    {
+      "epoch": 1.8498367791077257,
+      "grad_norm": 0.02612975798547268,
+      "learning_rate": 6.855903469733738e-05,
+      "loss": 0.06720860004425049,
+      "step": 1700
+    },
+    {
+      "epoch": 1.8607181719260066,
+      "grad_norm": 0.025018637999892235,
+      "learning_rate": 6.743883694527169e-05,
+      "loss": 0.06776891946792603,
+      "step": 1710
+    },
+    {
+      "epoch": 1.8715995647442871,
+      "grad_norm": 0.02438053861260414,
+      "learning_rate": 6.632319128997759e-05,
+      "loss": 0.07007836103439331,
+      "step": 1720
+    },
+    {
+      "epoch": 1.882480957562568,
+      "grad_norm": 0.02692045457661152,
+      "learning_rate": 6.52122537002998e-05,
+      "loss": 0.07095590829849244,
+      "step": 1730
+    },
+    {
+      "epoch": 1.8933623503808488,
+      "grad_norm": 0.031177183613181114,
+      "learning_rate": 6.410617948688892e-05,
+      "loss": 0.06855381727218628,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9042437431991295,
+      "grad_norm": 0.022553348913788795,
+      "learning_rate": 6.300512328048845e-05,
+      "loss": 0.07089903354644775,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9151251360174102,
+      "grad_norm": 0.024013947695493698,
+      "learning_rate": 6.190923901031757e-05,
+      "loss": 0.0708732008934021,
+      "step": 1760
+    },
+    {
+      "epoch": 1.926006528835691,
+      "grad_norm": 0.022148922085762024,
+      "learning_rate": 6.081867988255135e-05,
+      "loss": 0.07153573036193847,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9368879216539718,
+      "grad_norm": 0.047587718814611435,
+      "learning_rate": 5.973359835890246e-05,
+      "loss": 0.0697596788406372,
+      "step": 1780
+    },
+    {
+      "epoch": 1.9477693144722523,
+      "grad_norm": 0.02610265091061592,
+      "learning_rate": 5.865414613530669e-05,
+      "loss": 0.06653320789337158,
+      "step": 1790
+    },
+    {
+      "epoch": 1.9586507072905333,
+      "grad_norm": 0.02527604252099991,
+      "learning_rate": 5.758047412071568e-05,
+      "loss": 0.07076951861381531,
+      "step": 1800
+    },
+    {
+      "epoch": 1.9586507072905333,
+      "eval_loss": 0.07202806323766708,
+      "eval_runtime": 5.4058,
+      "eval_samples_per_second": 14.799,
+      "eval_steps_per_second": 1.85,
+      "step": 1800
+    },
+    {
+      "epoch": 1.969532100108814,
+      "grad_norm": 0.024923427030444145,
+      "learning_rate": 5.651273241599976e-05,
+      "loss": 0.06910277605056762,
+      "step": 1810
+    },
+    {
+      "epoch": 1.9804134929270947,
+      "grad_norm": 0.023715443909168243,
+      "learning_rate": 5.5451070292963434e-05,
+      "loss": 0.06721980571746826,
+      "step": 1820
+    },
+    {
+      "epoch": 1.9912948857453754,
+      "grad_norm": 0.019582392647862434,
+      "learning_rate": 5.439563617347711e-05,
+      "loss": 0.06764072179794312,
+      "step": 1830
+    },
+    {
+      "epoch": 2.002176278563656,
+      "grad_norm": 0.027950329706072807,
+      "learning_rate": 5.3346577608727463e-05,
+      "loss": 0.07124176025390624,
+      "step": 1840
+    },
+    {
+      "epoch": 2.013057671381937,
+      "grad_norm": 0.036231789737939835,
+      "learning_rate": 5.230404125858952e-05,
+      "loss": 0.05957789421081543,
+      "step": 1850
+    },
+    {
+      "epoch": 2.0239390642002175,
+      "grad_norm": 0.026198675855994225,
+      "learning_rate": 5.1268172871123535e-05,
+      "loss": 0.055424296855926515,
+      "step": 1860
+    },
+    {
+      "epoch": 2.0348204570184985,
+      "grad_norm": 0.025477459654211998,
+      "learning_rate": 5.023911726219902e-05,
+      "loss": 0.06114422678947449,
+      "step": 1870
+    },
+    {
+      "epoch": 2.045701849836779,
+      "grad_norm": 0.03552868962287903,
+      "learning_rate": 4.921701829524956e-05,
+      "loss": 0.058832204341888426,
+      "step": 1880
+    },
+    {
+      "epoch": 2.05658324265506,
+      "grad_norm": 0.03449210524559021,
+      "learning_rate": 4.820201886116022e-05,
+      "loss": 0.057713437080383304,
+      "step": 1890
+    },
+    {
+      "epoch": 2.067464635473341,
+      "grad_norm": 0.033065732568502426,
+      "learning_rate": 4.719426085829123e-05,
+      "loss": 0.05968807935714722,
+      "step": 1900
+    },
+    {
+      "epoch": 2.0783460282916213,
+      "grad_norm": 0.03002762608230114,
+      "learning_rate": 4.619388517264056e-05,
+      "loss": 0.060094785690307614,
+      "step": 1910
+    },
+    {
+      "epoch": 2.0892274211099022,
+      "grad_norm": 0.034517038613557816,
+      "learning_rate": 4.5201031658147766e-05,
+      "loss": 0.05601954460144043,
+      "step": 1920
+    },
+    {
+      "epoch": 2.1001088139281827,
+      "grad_norm": 0.03293369337916374,
+      "learning_rate": 4.4215839117142134e-05,
+      "loss": 0.06034954190254212,
+      "step": 1930
+    },
+    {
+      "epoch": 2.1109902067464636,
+      "grad_norm": 0.032126300036907196,
+      "learning_rate": 4.3238445280938154e-05,
+      "loss": 0.05510505437850952,
+      "step": 1940
+    },
+    {
+      "epoch": 2.121871599564744,
+      "grad_norm": 0.027177443727850914,
+      "learning_rate": 4.2268986790580125e-05,
+      "loss": 0.060755813121795656,
+      "step": 1950
+    },
+    {
+      "epoch": 2.132752992383025,
+      "grad_norm": 0.030124740675091743,
+      "learning_rate": 4.130759917773985e-05,
+      "loss": 0.06080647706985474,
+      "step": 1960
+    },
+    {
+      "epoch": 2.143634385201306,
+      "grad_norm": 0.03235238790512085,
+      "learning_rate": 4.0354416845769e-05,
+      "loss": 0.06485359072685241,
+      "step": 1970
+    },
+    {
+      "epoch": 2.1545157780195865,
+      "grad_norm": 0.033107586205005646,
+      "learning_rate": 3.940957305090935e-05,
+      "loss": 0.059920203685760495,
+      "step": 1980
+    },
+    {
+      "epoch": 2.1653971708378674,
+      "grad_norm": 0.02939060889184475,
+      "learning_rate": 3.8473199883663334e-05,
+      "loss": 0.0490723729133606,
+      "step": 1990
+    },
+    {
+      "epoch": 2.176278563656148,
+      "grad_norm": 0.03297962248325348,
+      "learning_rate": 3.754542825032766e-05,
+      "loss": 0.053455913066864015,
+      "step": 2000
+    },
+    {
+      "epoch": 2.176278563656148,
+      "eval_loss": 0.07033731788396835,
+      "eval_runtime": 5.4006,
+      "eval_samples_per_second": 14.813,
+      "eval_steps_per_second": 1.852,
+      "step": 2000
+    },
+    {
+      "epoch": 2.187159956474429,
+      "grad_norm": 0.026707574725151062,
+      "learning_rate": 3.6626387854692365e-05,
+      "loss": 0.061966001987457275,
+      "step": 2010
+    },
+    {
+      "epoch": 2.1980413492927093,
+      "grad_norm": 0.03455955162644386,
+      "learning_rate": 3.571620717990811e-05,
+      "loss": 0.05781112909317017,
+      "step": 2020
+    },
+    {
+      "epoch": 2.2089227421109903,
+      "grad_norm": 0.02576427161693573,
+      "learning_rate": 3.4815013470524004e-05,
+      "loss": 0.060006117820739745,
+      "step": 2030
+    },
+    {
+      "epoch": 2.2198041349292708,
+      "grad_norm": 0.03262777253985405,
+      "learning_rate": 3.392293271469866e-05,
+      "loss": 0.05959203839302063,
+      "step": 2040
+    },
+    {
+      "epoch": 2.2306855277475517,
+      "grad_norm": 0.031093895435333252,
+      "learning_rate": 3.304008962658677e-05,
+      "loss": 0.057662928104400636,
+      "step": 2050
+    },
+    {
+      "epoch": 2.2415669205658326,
+      "grad_norm": 0.030908960849046707,
+      "learning_rate": 3.216660762890413e-05,
+      "loss": 0.05794321894645691,
+      "step": 2060
+    },
+    {
+      "epoch": 2.252448313384113,
+      "grad_norm": 0.03053661808371544,
+      "learning_rate": 3.130260883567272e-05,
+      "loss": 0.05402252674102783,
+      "step": 2070
+    },
+    {
+      "epoch": 2.263329706202394,
+      "grad_norm": 0.03832336515188217,
+      "learning_rate": 3.0448214035149102e-05,
+      "loss": 0.06356992721557617,
+      "step": 2080
+    },
+    {
+      "epoch": 2.2742110990206745,
+      "grad_norm": 0.035567495971918106,
+      "learning_rate": 2.960354267293809e-05,
+      "loss": 0.057219624519348145,
+      "step": 2090
+    },
+    {
+      "epoch": 2.2850924918389555,
+      "grad_norm": 0.03574289754033089,
+      "learning_rate": 2.8768712835293922e-05,
+      "loss": 0.0654039740562439,
+      "step": 2100
+    },
+    {
+      "epoch": 2.295973884657236,
+      "grad_norm": 0.03358747810125351,
+      "learning_rate": 2.7943841232611812e-05,
+      "loss": 0.05935961604118347,
+      "step": 2110
+    },
+    {
+      "epoch": 2.306855277475517,
+      "grad_norm": 0.036444805562496185,
+      "learning_rate": 2.712904318311167e-05,
+      "loss": 0.059474557638168335,
+      "step": 2120
+    },
+    {
+      "epoch": 2.3177366702937974,
+      "grad_norm": 0.04165448620915413,
+      "learning_rate": 2.6324432596716276e-05,
+      "loss": 0.06448624134063721,
+      "step": 2130
+    },
+    {
+      "epoch": 2.3286180631120783,
+      "grad_norm": 0.03665981814265251,
+      "learning_rate": 2.5530121959126718e-05,
+      "loss": 0.055650746822357176,
+      "step": 2140
+    },
+    {
+      "epoch": 2.3394994559303592,
+      "grad_norm": 0.02873632311820984,
+      "learning_rate": 2.47462223160965e-05,
+      "loss": 0.05966170430183411,
+      "step": 2150
+    },
+    {
+      "epoch": 2.3503808487486397,
+      "grad_norm": 0.03279171139001846,
+      "learning_rate": 2.397284325790744e-05,
+      "loss": 0.06053815484046936,
+      "step": 2160
+    },
+    {
+      "epoch": 2.3612622415669207,
+      "grad_norm": 0.03719950094819069,
+      "learning_rate": 2.321009290404864e-05,
+      "loss": 0.0560091495513916,
+      "step": 2170
+    },
+    {
+      "epoch": 2.372143634385201,
+      "grad_norm": 0.03888888284564018,
+      "learning_rate": 2.2458077888101226e-05,
+      "loss": 0.058753132820129395,
+      "step": 2180
+    },
+    {
+      "epoch": 2.383025027203482,
+      "grad_norm": 0.0321931354701519,
+      "learning_rate": 2.1716903342830995e-05,
+      "loss": 0.05801019072532654,
+      "step": 2190
+    },
+    {
+      "epoch": 2.3939064200217626,
+      "grad_norm": 0.038627929985523224,
+      "learning_rate": 2.0986672885490378e-05,
+      "loss": 0.06117302179336548,
+      "step": 2200
+    },
+    {
+      "epoch": 2.3939064200217626,
+      "eval_loss": 0.0698942095041275,
+      "eval_runtime": 5.4996,
+      "eval_samples_per_second": 14.546,
+      "eval_steps_per_second": 1.818,
+      "step": 2200
+    },
+    {
+      "epoch": 2.4047878128400435,
+      "grad_norm": 0.03631049767136574,
+      "learning_rate": 2.0267488603332907e-05,
+      "loss": 0.05466359853744507,
+      "step": 2210
+    },
+    {
+      "epoch": 2.4156692056583244,
+      "grad_norm": 0.03237799555063248,
+      "learning_rate": 1.9559451039341147e-05,
+      "loss": 0.05731322765350342,
+      "step": 2220
+    },
+    {
+      "epoch": 2.426550598476605,
+      "grad_norm": 0.0349552221596241,
+      "learning_rate": 1.886265917817064e-05,
+      "loss": 0.06014159917831421,
+      "step": 2230
+    },
+    {
+      "epoch": 2.437431991294886,
+      "grad_norm": 0.0360136479139328,
+      "learning_rate": 1.8177210432311732e-05,
+      "loss": 0.05688718557357788,
+      "step": 2240
+    },
+    {
+      "epoch": 2.4483133841131663,
+      "grad_norm": 0.03673616796731949,
+      "learning_rate": 1.750320062847116e-05,
+      "loss": 0.058925163745880124,
+      "step": 2250
+    },
+    {
+      "epoch": 2.4591947769314473,
+      "grad_norm": 0.050142526626586914,
+      "learning_rate": 1.6840723994175356e-05,
+      "loss": 0.05712746381759644,
+      "step": 2260
+    },
+    {
+      "epoch": 2.4700761697497278,
+      "grad_norm": 0.03750143572688103,
+      "learning_rate": 1.6189873144597267e-05,
+      "loss": 0.056266939640045165,
+      "step": 2270
+    },
+    {
+      "epoch": 2.4809575625680087,
+      "grad_norm": 0.03403429314494133,
+      "learning_rate": 1.5550739069608676e-05,
+      "loss": 0.05650065541267395,
+      "step": 2280
+    },
+    {
+      "epoch": 2.4918389553862896,
+      "grad_norm": 0.030664529651403427,
+      "learning_rate": 1.492341112105966e-05,
+      "loss": 0.05500165820121765,
+      "step": 2290
+    },
+    {
+      "epoch": 2.50272034820457,
+      "grad_norm": 0.03341777250170708,
+      "learning_rate": 1.4307977000287064e-05,
+      "loss": 0.05969707369804382,
+      "step": 2300
+    },
+    {
+      "epoch": 2.513601741022851,
+      "grad_norm": 0.03309670463204384,
+      "learning_rate": 1.3704522745853743e-05,
+      "loss": 0.05166618824005127,
+      "step": 2310
+    },
+    {
+      "epoch": 2.5244831338411315,
+      "grad_norm": 0.03246740996837616,
+      "learning_rate": 1.311313272152025e-05,
+      "loss": 0.0567930817604065,
+      "step": 2320
+    },
+    {
+      "epoch": 2.5353645266594125,
+      "grad_norm": 0.036715373396873474,
+      "learning_rate": 1.2533889604450677e-05,
+      "loss": 0.05777972936630249,
+      "step": 2330
+    },
+    {
+      "epoch": 2.546245919477693,
+      "grad_norm": 0.03496614098548889,
+      "learning_rate": 1.196687437365427e-05,
+      "loss": 0.054594647884368894,
+      "step": 2340
+    },
+    {
+      "epoch": 2.557127312295974,
+      "grad_norm": 0.03273237124085426,
+      "learning_rate": 1.141216629866434e-05,
+      "loss": 0.05491857528686524,
+      "step": 2350
+    },
+    {
+      "epoch": 2.568008705114255,
+      "grad_norm": 0.03323022276163101,
+      "learning_rate": 1.0869842928456441e-05,
+      "loss": 0.05289023518562317,
+      "step": 2360
+    },
+    {
+      "epoch": 2.5788900979325353,
+      "grad_norm": 0.042828429490327835,
+      "learning_rate": 1.033998008060687e-05,
+      "loss": 0.05806971788406372,
+      "step": 2370
+    },
+    {
+      "epoch": 2.5897714907508163,
+      "grad_norm": 0.03340109810233116,
+      "learning_rate": 9.822651830693152e-06,
+      "loss": 0.05867668390274048,
+      "step": 2380
+    },
+    {
+      "epoch": 2.6006528835690967,
+      "grad_norm": 0.03182734549045563,
+      "learning_rate": 9.317930501938289e-06,
+      "loss": 0.05183411836624145,
+      "step": 2390
+    },
+    {
+      "epoch": 2.6115342763873777,
+      "grad_norm": 0.03279723972082138,
+      "learning_rate": 8.82588665509978e-06,
+      "loss": 0.06119837760925293,
+      "step": 2400
+    },
+    {
+      "epoch": 2.6115342763873777,
+      "eval_loss": 0.06727322936058044,
+      "eval_runtime": 5.4288,
+      "eval_samples_per_second": 14.736,
+      "eval_steps_per_second": 1.842,
+      "step": 2400
+    },
+    {
+      "epoch": 2.622415669205658,
+      "grad_norm": 0.03709235042333603,
+      "learning_rate": 8.346589078605215e-06,
+      "loss": 0.05637891292572021,
+      "step": 2410
+    },
+    {
+      "epoch": 2.633297062023939,
+      "grad_norm": 0.030506664887070656,
+      "learning_rate": 7.880104778935505e-06,
+      "loss": 0.05910442471504211,
+      "step": 2420
+    },
+    {
+      "epoch": 2.64417845484222,
+      "grad_norm": 0.03575789928436279,
+      "learning_rate": 7.426498971257278e-06,
+      "loss": 0.054590433835983276,
+      "step": 2430
+    },
+    {
+      "epoch": 2.6550598476605005,
+      "grad_norm": 0.035656314343214035,
+      "learning_rate": 6.985835070305768e-06,
+      "loss": 0.056293058395385745,
+      "step": 2440
+    },
+    {
+      "epoch": 2.6659412404787814,
+      "grad_norm": 0.03722505271434784,
+      "learning_rate": 6.5581746815192135e-06,
+      "loss": 0.058118653297424314,
+      "step": 2450
+    },
+    {
+      "epoch": 2.676822633297062,
+      "grad_norm": 0.035609811544418335,
+      "learning_rate": 6.143577592426453e-06,
+      "loss": 0.05446176528930664,
+      "step": 2460
+    },
+    {
+      "epoch": 2.687704026115343,
+      "grad_norm": 0.035058531910181046,
+      "learning_rate": 5.742101764288521e-06,
+      "loss": 0.05432036519050598,
+      "step": 2470
+    },
+    {
+      "epoch": 2.6985854189336234,
+      "grad_norm": 0.036867234855890274,
+      "learning_rate": 5.353803323995521e-06,
+      "loss": 0.05071129202842713,
+      "step": 2480
+    },
+    {
+      "epoch": 2.7094668117519043,
+      "grad_norm": 0.030480151996016502,
+      "learning_rate": 4.9787365562200985e-06,
+      "loss": 0.04942301511764526,
+      "step": 2490
+    },
+    {
+      "epoch": 2.7203482045701852,
+      "grad_norm": 0.0435580313205719,
+      "learning_rate": 4.616953895828269e-06,
+      "loss": 0.05862914323806763,
+      "step": 2500
+    },
+    {
+      "epoch": 2.7312295973884657,
+      "grad_norm": 0.03596170246601105,
+      "learning_rate": 4.268505920549016e-06,
+      "loss": 0.054695284366607665,
+      "step": 2510
+    },
+    {
+      "epoch": 2.742110990206746,
+      "grad_norm": 0.034583065658807755,
+      "learning_rate": 3.933441343903444e-06,
+      "loss": 0.04879617691040039,
+      "step": 2520
+    },
+    {
+      "epoch": 2.752992383025027,
+      "grad_norm": 0.03540559113025665,
+      "learning_rate": 3.61180700839453e-06,
+      "loss": 0.05482098460197449,
+      "step": 2530
+    },
+    {
+      "epoch": 2.763873775843308,
+      "grad_norm": 0.0387350358068943,
+      "learning_rate": 3.3036478789585044e-06,
+      "loss": 0.05537814497947693,
+      "step": 2540
+    },
+    {
+      "epoch": 2.7747551686615886,
+      "grad_norm": 0.03608511760830879,
+      "learning_rate": 3.0090070366786792e-06,
+      "loss": 0.05486249327659607,
+      "step": 2550
+    },
+    {
+      "epoch": 2.7856365614798695,
+      "grad_norm": 0.03293207287788391,
+      "learning_rate": 2.7279256727626546e-06,
+      "loss": 0.054778552055358885,
+      "step": 2560
+    },
+    {
+      "epoch": 2.7965179542981504,
+      "grad_norm": 0.03599146753549576,
+      "learning_rate": 2.4604430827837143e-06,
+      "loss": 0.05409228205680847,
+      "step": 2570
+    },
+    {
+      "epoch": 2.807399347116431,
+      "grad_norm": 0.03957504406571388,
+      "learning_rate": 2.2065966611872657e-06,
+      "loss": 0.06026080846786499,
+      "step": 2580
+    },
+    {
+      "epoch": 2.8182807399347114,
+      "grad_norm": 0.033959709107875824,
+      "learning_rate": 1.9664218960630533e-06,
+      "loss": 0.05407561063766479,
+      "step": 2590
+    },
+    {
+      "epoch": 2.8291621327529923,
+      "grad_norm": 0.04043720290064812,
+      "learning_rate": 1.7399523641838278e-06,
+      "loss": 0.0522362232208252,
+      "step": 2600
+    },
+    {
+      "epoch": 2.8291621327529923,
+      "eval_loss": 0.06709529459476471,
+      "eval_runtime": 5.6005,
+      "eval_samples_per_second": 14.284,
+      "eval_steps_per_second": 1.786,
+      "step": 2600
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2757,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.225807413317468e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

qwen3_8b_sft_rtl/checkpoint-2600/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63e1337140001fd786f5bfde65ea532ba76972f9e8e0826ca7b80b663dbc5302
+size 5176

qwen3_8b_sft_rtl/checkpoint-2757/README.md

@@ -0,0 +1,208 @@
+---
+base_model: /workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218
+library_name: peft
+pipeline_tag: text-generation
+tags:
+- base_model:adapter:/workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218
+- llama-factory
+- lora
+- transformers
+---
+
+# Model Card for Model ID
+
+<!-- Provide a quick summary of what the model is/does. -->
+
+
+
+## Model Details
+
+### Model Description
+
+<!-- Provide a longer summary of what this model is. -->
+
+
+
+- **Developed by:** [More Information Needed]
+- **Funded by [optional]:** [More Information Needed]
+- **Shared by [optional]:** [More Information Needed]
+- **Model type:** [More Information Needed]
+- **Language(s) (NLP):** [More Information Needed]
+- **License:** [More Information Needed]
+- **Finetuned from model [optional]:** [More Information Needed]
+
+### Model Sources [optional]
+
+<!-- Provide the basic links for the model. -->
+
+- **Repository:** [More Information Needed]
+- **Paper [optional]:** [More Information Needed]
+- **Demo [optional]:** [More Information Needed]
+
+## Uses
+
+<!-- Address questions around how the model is intended to be used, including the foreseeable users of the model and those affected by the model. -->
+
+### Direct Use
+
+<!-- This section is for the model use without fine-tuning or plugging into a larger ecosystem/app. -->
+
+[More Information Needed]
+
+### Downstream Use [optional]
+
+<!-- This section is for the model use when fine-tuned for a task, or when plugged into a larger ecosystem/app -->
+
+[More Information Needed]
+
+### Out-of-Scope Use
+
+<!-- This section addresses misuse, malicious use, and uses that the model will not work well for. -->
+
+[More Information Needed]
+
+## Bias, Risks, and Limitations
+
+<!-- This section is meant to convey both technical and sociotechnical limitations. -->
+
+[More Information Needed]
+
+### Recommendations
+
+<!-- This section is meant to convey recommendations with respect to the bias, risk, and technical limitations. -->
+
+Users (both direct and downstream) should be made aware of the risks, biases and limitations of the model. More information needed for further recommendations.
+
+## How to Get Started with the Model
+
+Use the code below to get started with the model.
+
+[More Information Needed]
+
+## Training Details
+
+### Training Data
+
+<!-- This should link to a Dataset Card, perhaps with a short stub of information on what the training data is all about as well as documentation related to data pre-processing or additional filtering. -->
+
+[More Information Needed]
+
+### Training Procedure
+
+<!-- This relates heavily to the Technical Specifications. Content here should link to that section when it is relevant to the training procedure. -->
+
+#### Preprocessing [optional]
+
+[More Information Needed]
+
+
+#### Training Hyperparameters
+
+- **Training regime:** [More Information Needed] <!--fp32, fp16 mixed precision, bf16 mixed precision, bf16 non-mixed precision, fp16 non-mixed precision, fp8 mixed precision -->
+
+#### Speeds, Sizes, Times [optional]
+
+<!-- This section provides information about throughput, start/end time, checkpoint size if relevant, etc. -->
+
+[More Information Needed]
+
+## Evaluation
+
+<!-- This section describes the evaluation protocols and provides the results. -->
+
+### Testing Data, Factors & Metrics
+
+#### Testing Data
+
+<!-- This should link to a Dataset Card if possible. -->
+
+[More Information Needed]
+
+#### Factors
+
+<!-- These are the things the evaluation is disaggregating by, e.g., subpopulations or domains. -->
+
+[More Information Needed]
+
+#### Metrics
+
+<!-- These are the evaluation metrics being used, ideally with a description of why. -->
+
+[More Information Needed]
+
+### Results
+
+[More Information Needed]
+
+#### Summary
+
+
+
+## Model Examination [optional]
+
+<!-- Relevant interpretability work for the model goes here -->
+
+[More Information Needed]
+
+## Environmental Impact
+
+<!-- Total emissions (in grams of CO2eq) and additional considerations, such as electricity usage, go here. Edit the suggested text below accordingly -->
+
+Carbon emissions can be estimated using the [Machine Learning Impact calculator](https://mlco2.github.io/impact#compute) presented in [Lacoste et al. (2019)](https://arxiv.org/abs/1910.09700).
+
+- **Hardware Type:** [More Information Needed]
+- **Hours used:** [More Information Needed]
+- **Cloud Provider:** [More Information Needed]
+- **Compute Region:** [More Information Needed]
+- **Carbon Emitted:** [More Information Needed]
+
+## Technical Specifications [optional]
+
+### Model Architecture and Objective
+
+[More Information Needed]
+
+### Compute Infrastructure
+
+[More Information Needed]
+
+#### Hardware
+
+[More Information Needed]
+
+#### Software
+
+[More Information Needed]
+
+## Citation [optional]
+
+<!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
+
+**BibTeX:**
+
+[More Information Needed]
+
+**APA:**
+
+[More Information Needed]
+
+## Glossary [optional]
+
+<!-- If relevant, include terms and calculations in this section that can help readers understand the model or model card. -->
+
+[More Information Needed]
+
+## More Information [optional]
+
+[More Information Needed]
+
+## Model Card Authors [optional]
+
+[More Information Needed]
+
+## Model Card Contact
+
+[More Information Needed]
+### Framework versions
+
+- PEFT 0.18.1

qwen3_8b_sft_rtl/checkpoint-2757/adapter_config.json

@@ -0,0 +1,46 @@
+{
+  "alora_invocation_tokens": null,
+  "alpha_pattern": {},
+  "arrow_config": null,
+  "auto_mapping": null,
+  "base_model_name_or_path": "/workspace/hf_cache/models--Qwen--Qwen3-8B/snapshots/b968826d9c46dd6066d109eabc6255188de91218",
+  "bias": "none",
+  "corda_config": null,
+  "ensure_weight_tying": false,
+  "eva_config": null,
+  "exclude_modules": null,
+  "fan_in_fan_out": false,
+  "inference_mode": true,
+  "init_lora_weights": true,
+  "layer_replication": null,
+  "layers_pattern": null,
+  "layers_to_transform": null,
+  "loftq_config": {},
+  "lora_alpha": 512,
+  "lora_bias": false,
+  "lora_dropout": 0.05,
+  "megatron_config": null,
+  "megatron_core": "megatron.core",
+  "modules_to_save": null,
+  "peft_type": "LORA",
+  "peft_version": "0.18.1",
+  "qalora_group_size": 16,
+  "r": 256,
+  "rank_pattern": {},
+  "revision": null,
+  "target_modules": [
+    "v_proj",
+    "down_proj",
+    "up_proj",
+    "gate_proj",
+    "q_proj",
+    "o_proj",
+    "k_proj"
+  ],
+  "target_parameters": null,
+  "task_type": "CAUSAL_LM",
+  "trainable_token_indices": null,
+  "use_dora": false,
+  "use_qalora": false,
+  "use_rslora": false
+}

qwen3_8b_sft_rtl/checkpoint-2757/adapter_model.safetensors

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d1af14129d5c06582bdfe78e4509c38e3850eebc975639aa610959c937020540
+size 2793475840

qwen3_8b_sft_rtl/checkpoint-2757/chat_template.jinja

@@ -0,0 +1,89 @@
+{%- if tools %}
+    {{- '<|im_start|>system\n' }}
+    {%- if messages[0].role == 'system' %}
+        {{- messages[0].content + '\n\n' }}
+    {%- endif %}
+    {{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
+    {%- for tool in tools %}
+        {{- "\n" }}
+        {{- tool | tojson }}
+    {%- endfor %}
+    {{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
+{%- else %}
+    {%- if messages[0].role == 'system' %}
+        {{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
+    {%- endif %}
+{%- endif %}
+{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
+{%- for message in messages[::-1] %}
+    {%- set index = (messages|length - 1) - loop.index0 %}
+    {%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
+        {%- set ns.multi_step_tool = false %}
+        {%- set ns.last_query_index = index %}
+    {%- endif %}
+{%- endfor %}
+{%- for message in messages %}
+    {%- if message.content is string %}
+        {%- set content = message.content %}
+    {%- else %}
+        {%- set content = '' %}
+    {%- endif %}
+    {%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
+        {{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
+    {%- elif message.role == "assistant" %}
+        {%- set reasoning_content = '' %}
+        {%- if message.reasoning_content is string %}
+            {%- set reasoning_content = message.reasoning_content %}
+        {%- else %}
+            {%- if '</think>' in content %}
+                {%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
+                {%- set content = content.split('</think>')[-1].lstrip('\n') %}
+            {%- endif %}
+        {%- endif %}
+        {%- if loop.index0 > ns.last_query_index %}
+            {%- if loop.last or (not loop.last and reasoning_content) %}
+                {{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
+            {%- else %}
+                {{- '<|im_start|>' + message.role + '\n' + content }}
+            {%- endif %}
+        {%- else %}
+            {{- '<|im_start|>' + message.role + '\n' + content }}
+        {%- endif %}
+        {%- if message.tool_calls %}
+            {%- for tool_call in message.tool_calls %}
+                {%- if (loop.first and content) or (not loop.first) %}
+                    {{- '\n' }}
+                {%- endif %}
+                {%- if tool_call.function %}
+                    {%- set tool_call = tool_call.function %}
+                {%- endif %}
+                {{- '<tool_call>\n{"name": "' }}
+                {{- tool_call.name }}
+                {{- '", "arguments": ' }}
+                {%- if tool_call.arguments is string %}
+                    {{- tool_call.arguments }}
+                {%- else %}
+                    {{- tool_call.arguments | tojson }}
+                {%- endif %}
+                {{- '}\n</tool_call>' }}
+            {%- endfor %}
+        {%- endif %}
+        {{- '<|im_end|>\n' }}
+    {%- elif message.role == "tool" %}
+        {%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
+            {{- '<|im_start|>user' }}
+        {%- endif %}
+        {{- '\n<tool_response>\n' }}
+        {{- content }}
+        {{- '\n</tool_response>' }}
+        {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
+            {{- '<|im_end|>\n' }}
+        {%- endif %}
+    {%- endif %}
+{%- endfor %}
+{%- if add_generation_prompt %}
+    {{- '<|im_start|>assistant\n' }}
+    {%- if enable_thinking is defined and enable_thinking is false %}
+        {{- '<think>\n\n</think>\n\n' }}
+    {%- endif %}
+{%- endif %}

qwen3_8b_sft_rtl/checkpoint-2757/optimizer.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0258f7ee385e565de033bd60ce1e9e393dd738f0c11f4fc2200168f4baf0e868
+size 5587244546

qwen3_8b_sft_rtl/checkpoint-2757/rng_state_0.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:56170aa85ba98b31d7b9ad66e0fe1ccf1838529ade1cb53bc6bf015552e0ca27
+size 15024

qwen3_8b_sft_rtl/checkpoint-2757/rng_state_1.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:04d86b1a1043f5cf4f9528089ab7f312c4aae7bb753c0efe64c67e85eeadffdb
+size 15024

qwen3_8b_sft_rtl/checkpoint-2757/rng_state_2.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f83a611f839f12297eee3ae38e80dce3067ddd4e1919999716768e45b14235b7
+size 15024

qwen3_8b_sft_rtl/checkpoint-2757/rng_state_3.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:add9ced2607dedb8c21267557081687f0fd0fcf4ac48c9fd2267cde20577a03d
+size 15024

qwen3_8b_sft_rtl/checkpoint-2757/scheduler.pt

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7c07d0b264968a0419e59457a7ad049d50f0f36f5c3d0752d605d0ef1251ae77
+size 1064

qwen3_8b_sft_rtl/checkpoint-2757/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650

qwen3_8b_sft_rtl/checkpoint-2757/tokenizer_config.json

@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

qwen3_8b_sft_rtl/checkpoint-2757/trainer_state.json

@@ -0,0 +1,2063 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 200,
+  "global_step": 2757,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01088139281828074,
+      "grad_norm": 0.2948031425476074,
+      "learning_rate": 1.8e-05,
+      "loss": 0.741086769104004,
+      "step": 10
+    },
+    {
+      "epoch": 0.02176278563656148,
+      "grad_norm": 0.15493904054164886,
+      "learning_rate": 3.8e-05,
+      "loss": 0.3913477897644043,
+      "step": 20
+    },
+    {
+      "epoch": 0.03264417845484222,
+      "grad_norm": 0.136435404419899,
+      "learning_rate": 5.8e-05,
+      "loss": 0.22936077117919923,
+      "step": 30
+    },
+    {
+      "epoch": 0.04352557127312296,
+      "grad_norm": 0.09819520264863968,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 0.17587745189666748,
+      "step": 40
+    },
+    {
+      "epoch": 0.0544069640914037,
+      "grad_norm": 0.10766161978244781,
+      "learning_rate": 9.8e-05,
+      "loss": 0.14949668645858766,
+      "step": 50
+    },
+    {
+      "epoch": 0.06528835690968444,
+      "grad_norm": 0.08924636989831924,
+      "learning_rate": 0.000118,
+      "loss": 0.13353660106658935,
+      "step": 60
+    },
+    {
+      "epoch": 0.07616974972796518,
+      "grad_norm": 0.08617673069238663,
+      "learning_rate": 0.000138,
+      "loss": 0.120628023147583,
+      "step": 70
+    },
+    {
+      "epoch": 0.08705114254624592,
+      "grad_norm": 0.06150153651833534,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 0.11396801471710205,
+      "step": 80
+    },
+    {
+      "epoch": 0.09793253536452666,
+      "grad_norm": 0.06216437742114067,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 0.1074080228805542,
+      "step": 90
+    },
+    {
+      "epoch": 0.1088139281828074,
+      "grad_norm": 0.07203802466392517,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 0.10922808647155761,
+      "step": 100
+    },
+    {
+      "epoch": 0.11969532100108814,
+      "grad_norm": 0.04227524250745773,
+      "learning_rate": 0.00019999433803003197,
+      "loss": 0.10778391361236572,
+      "step": 110
+    },
+    {
+      "epoch": 0.1305767138193689,
+      "grad_norm": 0.06469517946243286,
+      "learning_rate": 0.00019997476661131597,
+      "loss": 0.10001739263534545,
+      "step": 120
+    },
+    {
+      "epoch": 0.14145810663764963,
+      "grad_norm": 0.04663902893662453,
+      "learning_rate": 0.0001999412185784487,
+      "loss": 0.10446863174438477,
+      "step": 130
+    },
+    {
+      "epoch": 0.15233949945593037,
+      "grad_norm": 0.03584141284227371,
+      "learning_rate": 0.0001998936986214928,
+      "loss": 0.10438344478607178,
+      "step": 140
+    },
+    {
+      "epoch": 0.1632208922742111,
+      "grad_norm": 0.03158590570092201,
+      "learning_rate": 0.00019983221338380563,
+      "loss": 0.09951521158218384,
+      "step": 150
+    },
+    {
+      "epoch": 0.17410228509249184,
+      "grad_norm": 0.02909119985997677,
+      "learning_rate": 0.0001997567714611106,
+      "loss": 0.09911214709281921,
+      "step": 160
+    },
+    {
+      "epoch": 0.18498367791077258,
+      "grad_norm": 0.030274737626314163,
+      "learning_rate": 0.0001996673834002955,
+      "loss": 0.09334277510643005,
+      "step": 170
+    },
+    {
+      "epoch": 0.19586507072905332,
+      "grad_norm": 0.028922295197844505,
+      "learning_rate": 0.00019956406169793805,
+      "loss": 0.09431872367858887,
+      "step": 180
+    },
+    {
+      "epoch": 0.20674646354733406,
+      "grad_norm": 0.030195271596312523,
+      "learning_rate": 0.00019944682079855867,
+      "loss": 0.10089273452758789,
+      "step": 190
+    },
+    {
+      "epoch": 0.2176278563656148,
+      "grad_norm": 0.029216868802905083,
+      "learning_rate": 0.0001993156770926014,
+      "loss": 0.0949325442314148,
+      "step": 200
+    },
+    {
+      "epoch": 0.2176278563656148,
+      "eval_loss": 0.09785260260105133,
+      "eval_runtime": 5.4576,
+      "eval_samples_per_second": 14.659,
+      "eval_steps_per_second": 1.832,
+      "step": 200
+    },
+    {
+      "epoch": 0.22850924918389554,
+      "grad_norm": 0.029826847836375237,
+      "learning_rate": 0.00019917064891414232,
+      "loss": 0.09331356287002564,
+      "step": 210
+    },
+    {
+      "epoch": 0.23939064200217627,
+      "grad_norm": 0.03143204376101494,
+      "learning_rate": 0.00019901175653832645,
+      "loss": 0.09654195308685302,
+      "step": 220
+    },
+    {
+      "epoch": 0.250272034820457,
+      "grad_norm": 0.02917400933802128,
+      "learning_rate": 0.00019883902217853326,
+      "loss": 0.09114633202552795,
+      "step": 230
+    },
+    {
+      "epoch": 0.2611534276387378,
+      "grad_norm": 0.033463530242443085,
+      "learning_rate": 0.0001986524699832712,
+      "loss": 0.09384369254112243,
+      "step": 240
+    },
+    {
+      "epoch": 0.2720348204570185,
+      "grad_norm": 0.027332467958331108,
+      "learning_rate": 0.00019845212603280174,
+      "loss": 0.08879860639572143,
+      "step": 250
+    },
+    {
+      "epoch": 0.28291621327529926,
+      "grad_norm": 0.03050387091934681,
+      "learning_rate": 0.00019823801833549327,
+      "loss": 0.09065196514129639,
+      "step": 260
+    },
+    {
+      "epoch": 0.29379760609357997,
+      "grad_norm": 0.024866925552487373,
+      "learning_rate": 0.00019801017682390552,
+      "loss": 0.09161882400512696,
+      "step": 270
+    },
+    {
+      "epoch": 0.30467899891186073,
+      "grad_norm": 0.037345923483371735,
+      "learning_rate": 0.0001977686333506049,
+      "loss": 0.09159514904022217,
+      "step": 280
+    },
+    {
+      "epoch": 0.31556039173014144,
+      "grad_norm": 0.0632031187415123,
+      "learning_rate": 0.00019751342168371154,
+      "loss": 0.0902558445930481,
+      "step": 290
+    },
+    {
+      "epoch": 0.3264417845484222,
+      "grad_norm": 0.029686614871025085,
+      "learning_rate": 0.00019724457750217839,
+      "loss": 0.09181964993476868,
+      "step": 300
+    },
+    {
+      "epoch": 0.3373231773667029,
+      "grad_norm": 0.03043133206665516,
+      "learning_rate": 0.0001969621383908033,
+      "loss": 0.08723740577697754,
+      "step": 310
+    },
+    {
+      "epoch": 0.3482045701849837,
+      "grad_norm": 0.041717205196619034,
+      "learning_rate": 0.00019666614383497457,
+      "loss": 0.09164115190505981,
+      "step": 320
+    },
+    {
+      "epoch": 0.3590859630032644,
+      "grad_norm": 0.03025348111987114,
+      "learning_rate": 0.00019635663521515087,
+      "loss": 0.08903414011001587,
+      "step": 330
+    },
+    {
+      "epoch": 0.36996735582154516,
+      "grad_norm": 0.030753403902053833,
+      "learning_rate": 0.00019603365580107615,
+      "loss": 0.08611454963684081,
+      "step": 340
+    },
+    {
+      "epoch": 0.3808487486398259,
+      "grad_norm": 0.021834490820765495,
+      "learning_rate": 0.0001956972507457306,
+      "loss": 0.08958455920219421,
+      "step": 350
+    },
+    {
+      "epoch": 0.39173014145810664,
+      "grad_norm": 0.025724470615386963,
+      "learning_rate": 0.000195347467079018,
+      "loss": 0.08454820513725281,
+      "step": 360
+    },
+    {
+      "epoch": 0.40261153427638735,
+      "grad_norm": 0.034213438630104065,
+      "learning_rate": 0.00019498435370119094,
+      "loss": 0.09067847728729247,
+      "step": 370
+    },
+    {
+      "epoch": 0.4134929270946681,
+      "grad_norm": 0.0437682569026947,
+      "learning_rate": 0.00019460796137601467,
+      "loss": 0.0873767077922821,
+      "step": 380
+    },
+    {
+      "epoch": 0.4243743199129489,
+      "grad_norm": 0.027341296896338463,
+      "learning_rate": 0.00019421834272367,
+      "loss": 0.08544692993164063,
+      "step": 390
+    },
+    {
+      "epoch": 0.4352557127312296,
+      "grad_norm": 0.030272383242845535,
+      "learning_rate": 0.00019381555221339707,
+      "loss": 0.08988077044487,
+      "step": 400
+    },
+    {
+      "epoch": 0.4352557127312296,
+      "eval_loss": 0.09069998562335968,
+      "eval_runtime": 5.3759,
+      "eval_samples_per_second": 14.881,
+      "eval_steps_per_second": 1.86,
+      "step": 400
+    },
+    {
+      "epoch": 0.44613710554951036,
+      "grad_norm": 0.030490044504404068,
+      "learning_rate": 0.00019339964615588043,
+      "loss": 0.088669753074646,
+      "step": 410
+    },
+    {
+      "epoch": 0.45701849836779107,
+      "grad_norm": 0.030257854610681534,
+      "learning_rate": 0.0001929706826953767,
+      "loss": 0.09294685125350952,
+      "step": 420
+    },
+    {
+      "epoch": 0.46789989118607184,
+      "grad_norm": 0.03422554209828377,
+      "learning_rate": 0.00019252872180158605,
+      "loss": 0.08597595095634461,
+      "step": 430
+    },
+    {
+      "epoch": 0.47878128400435255,
+      "grad_norm": 0.026077628135681152,
+      "learning_rate": 0.0001920738252612682,
+      "loss": 0.08706499338150024,
+      "step": 440
+    },
+    {
+      "epoch": 0.4896626768226333,
+      "grad_norm": 0.025868231430649757,
+      "learning_rate": 0.0001916060566696045,
+      "loss": 0.08425920009613037,
+      "step": 450
+    },
+    {
+      "epoch": 0.500544069640914,
+      "grad_norm": 0.03231549635529518,
+      "learning_rate": 0.00019112548142130757,
+      "loss": 0.08545957803726197,
+      "step": 460
+    },
+    {
+      "epoch": 0.5114254624591947,
+      "grad_norm": 0.03557935357093811,
+      "learning_rate": 0.00019063216670147854,
+      "loss": 0.08239468932151794,
+      "step": 470
+    },
+    {
+      "epoch": 0.5223068552774756,
+      "grad_norm": 0.023028500378131866,
+      "learning_rate": 0.00019012618147621488,
+      "loss": 0.08288769721984864,
+      "step": 480
+    },
+    {
+      "epoch": 0.5331882480957563,
+      "grad_norm": 0.021350380033254623,
+      "learning_rate": 0.00018960759648296865,
+      "loss": 0.08916928768157958,
+      "step": 490
+    },
+    {
+      "epoch": 0.544069640914037,
+      "grad_norm": 0.028527546674013138,
+      "learning_rate": 0.0001890764842206573,
+      "loss": 0.09119058847427368,
+      "step": 500
+    },
+    {
+      "epoch": 0.5549510337323177,
+      "grad_norm": 0.023469405248761177,
+      "learning_rate": 0.00018853291893952833,
+      "loss": 0.08246206045150757,
+      "step": 510
+    },
+    {
+      "epoch": 0.5658324265505985,
+      "grad_norm": 0.028494561091065407,
+      "learning_rate": 0.0001879769766307789,
+      "loss": 0.0803530752658844,
+      "step": 520
+    },
+    {
+      "epoch": 0.5767138193688792,
+      "grad_norm": 0.027940815314650536,
+      "learning_rate": 0.00018740873501593214,
+      "loss": 0.08848085403442382,
+      "step": 530
+    },
+    {
+      "epoch": 0.5875952121871599,
+      "grad_norm": 0.02026485837996006,
+      "learning_rate": 0.0001868282735359717,
+      "loss": 0.08710720539093017,
+      "step": 540
+    },
+    {
+      "epoch": 0.5984766050054406,
+      "grad_norm": 0.026024585589766502,
+      "learning_rate": 0.00018623567334023563,
+      "loss": 0.07982800006866456,
+      "step": 550
+    },
+    {
+      "epoch": 0.6093579978237215,
+      "grad_norm": 0.029224518686532974,
+      "learning_rate": 0.0001856310172750717,
+      "loss": 0.08492777943611145,
+      "step": 560
+    },
+    {
+      "epoch": 0.6202393906420022,
+      "grad_norm": 0.02598229981958866,
+      "learning_rate": 0.00018501438987225536,
+      "loss": 0.08290873765945435,
+      "step": 570
+    },
+    {
+      "epoch": 0.6311207834602829,
+      "grad_norm": 0.023201750591397285,
+      "learning_rate": 0.00018438587733717187,
+      "loss": 0.08711419701576233,
+      "step": 580
+    },
+    {
+      "epoch": 0.6420021762785637,
+      "grad_norm": 0.026234455406665802,
+      "learning_rate": 0.00018374556753676506,
+      "loss": 0.08342958688735962,
+      "step": 590
+    },
+    {
+      "epoch": 0.6528835690968444,
+      "grad_norm": 0.021891364827752113,
+      "learning_rate": 0.00018309354998725306,
+      "loss": 0.07860854864120484,
+      "step": 600
+    },
+    {
+      "epoch": 0.6528835690968444,
+      "eval_loss": 0.08754301071166992,
+      "eval_runtime": 5.5341,
+      "eval_samples_per_second": 14.456,
+      "eval_steps_per_second": 1.807,
+      "step": 600
+    },
+    {
+      "epoch": 0.6637649619151251,
+      "grad_norm": 0.026508808135986328,
+      "learning_rate": 0.00018242991584161393,
+      "loss": 0.08734351992607117,
+      "step": 610
+    },
+    {
+      "epoch": 0.6746463547334058,
+      "grad_norm": 0.02286313846707344,
+      "learning_rate": 0.0001817547578768424,
+      "loss": 0.08437882661819458,
+      "step": 620
+    },
+    {
+      "epoch": 0.6855277475516867,
+      "grad_norm": 0.02370450273156166,
+      "learning_rate": 0.00018106817048097937,
+      "loss": 0.08016583323478699,
+      "step": 630
+    },
+    {
+      "epoch": 0.6964091403699674,
+      "grad_norm": 0.025123964995145798,
+      "learning_rate": 0.0001803702496399165,
+      "loss": 0.076961088180542,
+      "step": 640
+    },
+    {
+      "epoch": 0.7072905331882481,
+      "grad_norm": 0.019611909985542297,
+      "learning_rate": 0.0001796610929239771,
+      "loss": 0.08545655012130737,
+      "step": 650
+    },
+    {
+      "epoch": 0.7181719260065288,
+      "grad_norm": 0.020668957382440567,
+      "learning_rate": 0.00017894079947427573,
+      "loss": 0.0809054434299469,
+      "step": 660
+    },
+    {
+      "epoch": 0.7290533188248096,
+      "grad_norm": 0.028357358649373055,
+      "learning_rate": 0.0001782094699888582,
+      "loss": 0.07935836911201477,
+      "step": 670
+    },
+    {
+      "epoch": 0.7399347116430903,
+      "grad_norm": 0.018039165064692497,
+      "learning_rate": 0.00017746720670862363,
+      "loss": 0.08474135994911194,
+      "step": 680
+    },
+    {
+      "epoch": 0.750816104461371,
+      "grad_norm": 0.02181115373969078,
+      "learning_rate": 0.0001767141134030313,
+      "loss": 0.08088613748550415,
+      "step": 690
+    },
+    {
+      "epoch": 0.7616974972796517,
+      "grad_norm": 0.029043374583125114,
+      "learning_rate": 0.00017595029535559336,
+      "loss": 0.078003191947937,
+      "step": 700
+    },
+    {
+      "epoch": 0.7725788900979326,
+      "grad_norm": 0.024250885471701622,
+      "learning_rate": 0.00017517585934915595,
+      "loss": 0.08291269540786743,
+      "step": 710
+    },
+    {
+      "epoch": 0.7834602829162133,
+      "grad_norm": 0.028013000264763832,
+      "learning_rate": 0.00017439091365097106,
+      "loss": 0.08375194072723388,
+      "step": 720
+    },
+    {
+      "epoch": 0.794341675734494,
+      "grad_norm": 0.030651913955807686,
+      "learning_rate": 0.00017359556799756052,
+      "loss": 0.08069210052490235,
+      "step": 730
+    },
+    {
+      "epoch": 0.8052230685527747,
+      "grad_norm": 0.028212785720825195,
+      "learning_rate": 0.00017278993357937455,
+      "loss": 0.08342626094818115,
+      "step": 740
+    },
+    {
+      "epoch": 0.8161044613710555,
+      "grad_norm": 0.024551352486014366,
+      "learning_rate": 0.0001719741230252473,
+      "loss": 0.0814591646194458,
+      "step": 750
+    },
+    {
+      "epoch": 0.8269858541893362,
+      "grad_norm": 0.022670986130833626,
+      "learning_rate": 0.00017114825038665123,
+      "loss": 0.08308478593826293,
+      "step": 760
+    },
+    {
+      "epoch": 0.8378672470076169,
+      "grad_norm": 0.03288818150758743,
+      "learning_rate": 0.0001703124311217524,
+      "loss": 0.07733856439590454,
+      "step": 770
+    },
+    {
+      "epoch": 0.8487486398258978,
+      "grad_norm": 0.02053298056125641,
+      "learning_rate": 0.00016946678207926938,
+      "loss": 0.0828587532043457,
+      "step": 780
+    },
+    {
+      "epoch": 0.8596300326441785,
+      "grad_norm": 0.024455999955534935,
+      "learning_rate": 0.0001686114214821375,
+      "loss": 0.07782068252563476,
+      "step": 790
+    },
+    {
+      "epoch": 0.8705114254624592,
+      "grad_norm": 0.022575655952095985,
+      "learning_rate": 0.00016774646891098134,
+      "loss": 0.08273504972457886,
+      "step": 800
+    },
+    {
+      "epoch": 0.8705114254624592,
+      "eval_loss": 0.08349094539880753,
+      "eval_runtime": 5.5064,
+      "eval_samples_per_second": 14.529,
+      "eval_steps_per_second": 1.816,
+      "step": 800
+    },
+    {
+      "epoch": 0.8813928182807399,
+      "grad_norm": 0.02310466580092907,
+      "learning_rate": 0.00016687204528739694,
+      "loss": 0.08457955718040466,
+      "step": 810
+    },
+    {
+      "epoch": 0.8922742110990207,
+      "grad_norm": 0.02504105120897293,
+      "learning_rate": 0.00016598827285704696,
+      "loss": 0.08319224119186401,
+      "step": 820
+    },
+    {
+      "epoch": 0.9031556039173014,
+      "grad_norm": 0.02492046356201172,
+      "learning_rate": 0.00016509527517257042,
+      "loss": 0.08071508407592773,
+      "step": 830
+    },
+    {
+      "epoch": 0.9140369967355821,
+      "grad_norm": 0.023182213306427002,
+      "learning_rate": 0.0001641931770763101,
+      "loss": 0.08242483139038086,
+      "step": 840
+    },
+    {
+      "epoch": 0.9249183895538629,
+      "grad_norm": 0.022171733900904655,
+      "learning_rate": 0.00016328210468285911,
+      "loss": 0.08000438213348389,
+      "step": 850
+    },
+    {
+      "epoch": 0.9357997823721437,
+      "grad_norm": 0.023160789161920547,
+      "learning_rate": 0.00016236218536143013,
+      "loss": 0.08225069642066955,
+      "step": 860
+    },
+    {
+      "epoch": 0.9466811751904244,
+      "grad_norm": 0.021988507360219955,
+      "learning_rate": 0.0001614335477180488,
+      "loss": 0.07968268394470215,
+      "step": 870
+    },
+    {
+      "epoch": 0.9575625680087051,
+      "grad_norm": 0.0209442637860775,
+      "learning_rate": 0.00016049632157757473,
+      "loss": 0.08057799339294433,
+      "step": 880
+    },
+    {
+      "epoch": 0.9684439608269858,
+      "grad_norm": 0.02684466354548931,
+      "learning_rate": 0.00015955063796555137,
+      "loss": 0.07912254929542542,
+      "step": 890
+    },
+    {
+      "epoch": 0.9793253536452666,
+      "grad_norm": 0.022005856037139893,
+      "learning_rate": 0.00015859662908988885,
+      "loss": 0.07660449743270874,
+      "step": 900
+    },
+    {
+      "epoch": 0.9902067464635473,
+      "grad_norm": 0.022977208718657494,
+      "learning_rate": 0.0001576344283223809,
+      "loss": 0.07685725092887878,
+      "step": 910
+    },
+    {
+      "epoch": 1.001088139281828,
+      "grad_norm": 0.020924242213368416,
+      "learning_rate": 0.00015666417018005938,
+      "loss": 0.07933401465415954,
+      "step": 920
+    },
+    {
+      "epoch": 1.0119695321001088,
+      "grad_norm": 0.02789444290101528,
+      "learning_rate": 0.00015568599030638854,
+      "loss": 0.07339736819267273,
+      "step": 930
+    },
+    {
+      "epoch": 1.0228509249183895,
+      "grad_norm": 0.02638629637658596,
+      "learning_rate": 0.00015470002545230193,
+      "loss": 0.07634537220001221,
+      "step": 940
+    },
+    {
+      "epoch": 1.0337323177366704,
+      "grad_norm": 0.029259871691465378,
+      "learning_rate": 0.0001537064134570844,
+      "loss": 0.08004393577575683,
+      "step": 950
+    },
+    {
+      "epoch": 1.0446137105549511,
+      "grad_norm": 0.03044624626636505,
+      "learning_rate": 0.0001527052932291019,
+      "loss": 0.07677503824234008,
+      "step": 960
+    },
+    {
+      "epoch": 1.0554951033732318,
+      "grad_norm": 0.02535896934568882,
+      "learning_rate": 0.00015169680472638218,
+      "loss": 0.07567052841186524,
+      "step": 970
+    },
+    {
+      "epoch": 1.0663764961915125,
+      "grad_norm": 0.02594912238419056,
+      "learning_rate": 0.00015068108893704797,
+      "loss": 0.0768008291721344,
+      "step": 980
+    },
+    {
+      "epoch": 1.0772578890097932,
+      "grad_norm": 0.026760386303067207,
+      "learning_rate": 0.0001496582878596071,
+      "loss": 0.07604062557220459,
+      "step": 990
+    },
+    {
+      "epoch": 1.088139281828074,
+      "grad_norm": 0.02427523396909237,
+      "learning_rate": 0.00014862854448310074,
+      "loss": 0.07631718516349792,
+      "step": 1000
+    },
+    {
+      "epoch": 1.088139281828074,
+      "eval_loss": 0.08224387466907501,
+      "eval_runtime": 5.3767,
+      "eval_samples_per_second": 14.879,
+      "eval_steps_per_second": 1.86,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0990206746463547,
+      "grad_norm": 0.023537471890449524,
+      "learning_rate": 0.00014759200276711316,
+      "loss": 0.0752741813659668,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1099020674646354,
+      "grad_norm": 0.028346385806798935,
+      "learning_rate": 0.00014654880762164624,
+      "loss": 0.07320383191108704,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1207834602829163,
+      "grad_norm": 0.024628447368741035,
+      "learning_rate": 0.00014549910488686078,
+      "loss": 0.08079155683517455,
+      "step": 1030
+    },
+    {
+      "epoch": 1.131664853101197,
+      "grad_norm": 0.024834809824824333,
+      "learning_rate": 0.0001444430413126877,
+      "loss": 0.07409957647323609,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1425462459194777,
+      "grad_norm": 0.03418092057108879,
+      "learning_rate": 0.0001433807645383123,
+      "loss": 0.07469584345817566,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1534276387377584,
+      "grad_norm": 0.022386226803064346,
+      "learning_rate": 0.00014231242307153417,
+      "loss": 0.07519133687019348,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1643090315560392,
+      "grad_norm": 0.023596055805683136,
+      "learning_rate": 0.00014123816626800535,
+      "loss": 0.07806066274642945,
+      "step": 1070
+    },
+    {
+      "epoch": 1.1751904243743199,
+      "grad_norm": 0.025627054274082184,
+      "learning_rate": 0.00014015814431035046,
+      "loss": 0.07402833104133606,
+      "step": 1080
+    },
+    {
+      "epoch": 1.1860718171926006,
+      "grad_norm": 0.021048523485660553,
+      "learning_rate": 0.00013907250818717082,
+      "loss": 0.07758893370628357,
+      "step": 1090
+    },
+    {
+      "epoch": 1.1969532100108813,
+      "grad_norm": 0.028320247307419777,
+      "learning_rate": 0.00013798140967193607,
+      "loss": 0.0765085220336914,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2078346028291622,
+      "grad_norm": 0.027219558134675026,
+      "learning_rate": 0.0001368850013017659,
+      "loss": 0.07446168661117554,
+      "step": 1110
+    },
+    {
+      "epoch": 1.218715995647443,
+      "grad_norm": 0.022490857169032097,
+      "learning_rate": 0.0001357834363561054,
+      "loss": 0.07697376012802123,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2295973884657236,
+      "grad_norm": 0.02330136112868786,
+      "learning_rate": 0.00013467686883529616,
+      "loss": 0.07688432335853576,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2404787812840044,
+      "grad_norm": 0.025878841057419777,
+      "learning_rate": 0.0001335654534390468,
+      "loss": 0.07023751735687256,
+      "step": 1140
+    },
+    {
+      "epoch": 1.251360174102285,
+      "grad_norm": 0.023564936593174934,
+      "learning_rate": 0.0001324493455448059,
+      "loss": 0.0761339783668518,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2622415669205658,
+      "grad_norm": 0.01974901743233204,
+      "learning_rate": 0.0001313287011860399,
+      "loss": 0.07061100006103516,
+      "step": 1160
+    },
+    {
+      "epoch": 1.2731229597388465,
+      "grad_norm": 0.028187213465571404,
+      "learning_rate": 0.00013020367703041934,
+      "loss": 0.07600026726722717,
+      "step": 1170
+    },
+    {
+      "epoch": 1.2840043525571274,
+      "grad_norm": 0.027759388089179993,
+      "learning_rate": 0.0001290744303579167,
+      "loss": 0.07545133829116821,
+      "step": 1180
+    },
+    {
+      "epoch": 1.2948857453754081,
+      "grad_norm": 0.03569910302758217,
+      "learning_rate": 0.00012794111903881818,
+      "loss": 0.07300611138343811,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3057671381936888,
+      "grad_norm": 0.026020081713795662,
+      "learning_rate": 0.00012680390151165337,
+      "loss": 0.07233045101165772,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3057671381936888,
+      "eval_loss": 0.08009672164916992,
+      "eval_runtime": 5.448,
+      "eval_samples_per_second": 14.684,
+      "eval_steps_per_second": 1.836,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3166485310119695,
+      "grad_norm": 0.02722887694835663,
+      "learning_rate": 0.0001256629367610454,
+      "loss": 0.0722274899482727,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3275299238302503,
+      "grad_norm": 0.021629413589835167,
+      "learning_rate": 0.0001245183842954844,
+      "loss": 0.07423741221427918,
+      "step": 1220
+    },
+    {
+      "epoch": 1.338411316648531,
+      "grad_norm": 0.030425263568758965,
+      "learning_rate": 0.00012337040412502823,
+      "loss": 0.07771967649459839,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3492927094668117,
+      "grad_norm": 0.022577406838536263,
+      "learning_rate": 0.00012221915673893267,
+      "loss": 0.07685770392417908,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3601741022850926,
+      "grad_norm": 0.026748720556497574,
+      "learning_rate": 0.00012106480308321498,
+      "loss": 0.0744680941104889,
+      "step": 1250
+    },
+    {
+      "epoch": 1.371055495103373,
+      "grad_norm": 0.024764206260442734,
+      "learning_rate": 0.00011990750453815324,
+      "loss": 0.07823204398155212,
+      "step": 1260
+    },
+    {
+      "epoch": 1.381936887921654,
+      "grad_norm": 0.023767156526446342,
+      "learning_rate": 0.00011874742289572531,
+      "loss": 0.06924179792404175,
+      "step": 1270
+    },
+    {
+      "epoch": 1.3928182807399347,
+      "grad_norm": 0.025964556261897087,
+      "learning_rate": 0.00011758472033698993,
+      "loss": 0.06726616621017456,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4036996735582155,
+      "grad_norm": 0.024118445813655853,
+      "learning_rate": 0.00011641955940941374,
+      "loss": 0.07383161187171935,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4145810663764962,
+      "grad_norm": 0.0229018721729517,
+      "learning_rate": 0.00011525210300414673,
+      "loss": 0.07341222763061524,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4254624591947769,
+      "grad_norm": 0.02409880980849266,
+      "learning_rate": 0.00011408251433325006,
+      "loss": 0.07354844808578491,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4363438520130578,
+      "grad_norm": 0.039964865893125534,
+      "learning_rate": 0.00011291095690687847,
+      "loss": 0.07191305160522461,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4472252448313383,
+      "grad_norm": 0.02640443667769432,
+      "learning_rate": 0.00011173759451042168,
+      "loss": 0.07049686908721924,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4581066376496192,
+      "grad_norm": 0.022137323394417763,
+      "learning_rate": 0.0001105625911816067,
+      "loss": 0.07753136157989501,
+      "step": 1340
+    },
+    {
+      "epoch": 1.4689880304679,
+      "grad_norm": 0.02352975122630596,
+      "learning_rate": 0.0001093861111875652,
+      "loss": 0.07463480234146118,
+      "step": 1350
+    },
+    {
+      "epoch": 1.4798694232861807,
+      "grad_norm": 0.02214525081217289,
+      "learning_rate": 0.0001082083190018688,
+      "loss": 0.06845179796218873,
+      "step": 1360
+    },
+    {
+      "epoch": 1.4907508161044614,
+      "grad_norm": 0.025555888190865517,
+      "learning_rate": 0.00010702937928153533,
+      "loss": 0.07124170064926147,
+      "step": 1370
+    },
+    {
+      "epoch": 1.501632208922742,
+      "grad_norm": 0.02631489560008049,
+      "learning_rate": 0.00010584945684400963,
+      "loss": 0.07157056331634522,
+      "step": 1380
+    },
+    {
+      "epoch": 1.512513601741023,
+      "grad_norm": 0.026430578902363777,
+      "learning_rate": 0.0001046687166441217,
+      "loss": 0.07220081686973571,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5233949945593035,
+      "grad_norm": 0.024772843345999718,
+      "learning_rate": 0.00010348732375102601,
+      "loss": 0.07243863940238952,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5233949945593035,
+      "eval_loss": 0.07733941823244095,
+      "eval_runtime": 5.3743,
+      "eval_samples_per_second": 14.886,
+      "eval_steps_per_second": 1.861,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5342763873775844,
+      "grad_norm": 0.0237529706209898,
+      "learning_rate": 0.00010230544332512442,
+      "loss": 0.06972206830978393,
+      "step": 1410
+    },
+    {
+      "epoch": 1.545157780195865,
+      "grad_norm": 0.02250276878476143,
+      "learning_rate": 0.00010112324059497649,
+      "loss": 0.07575106620788574,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5560391730141458,
+      "grad_norm": 0.02346464991569519,
+      "learning_rate": 9.994088083420044e-05,
+      "loss": 0.07188562154769898,
+      "step": 1430
+    },
+    {
+      "epoch": 1.5669205658324266,
+      "grad_norm": 0.023935239762067795,
+      "learning_rate": 9.875852933836765e-05,
+      "loss": 0.0751798152923584,
+      "step": 1440
+    },
+    {
+      "epoch": 1.5778019586507073,
+      "grad_norm": 0.02204659767448902,
+      "learning_rate": 9.757635140189394e-05,
+      "loss": 0.06870698928833008,
+      "step": 1450
+    },
+    {
+      "epoch": 1.588683351468988,
+      "grad_norm": 0.022586360573768616,
+      "learning_rate": 9.639451229493133e-05,
+      "loss": 0.07310816049575805,
+      "step": 1460
+    },
+    {
+      "epoch": 1.5995647442872687,
+      "grad_norm": 0.02241821214556694,
+      "learning_rate": 9.521317724026299e-05,
+      "loss": 0.07275789976119995,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6104461371055496,
+      "grad_norm": 0.019981661811470985,
+      "learning_rate": 9.403251139020486e-05,
+      "loss": 0.07513962984085083,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6213275299238301,
+      "grad_norm": 0.0203393641859293,
+      "learning_rate": 9.285267980351719e-05,
+      "loss": 0.0734050989151001,
+      "step": 1490
+    },
+    {
+      "epoch": 1.632208922742111,
+      "grad_norm": 0.027680471539497375,
+      "learning_rate": 9.167384742232898e-05,
+      "loss": 0.06607317924499512,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6430903155603918,
+      "grad_norm": 0.026372363790869713,
+      "learning_rate": 9.049617904907877e-05,
+      "loss": 0.07146948575973511,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6539717083786725,
+      "grad_norm": 0.02281905524432659,
+      "learning_rate": 8.931983932347533e-05,
+      "loss": 0.07334067225456238,
+      "step": 1520
+    },
+    {
+      "epoch": 1.6648531011969532,
+      "grad_norm": 0.02359774336218834,
+      "learning_rate": 8.814499269948053e-05,
+      "loss": 0.0697287678718567,
+      "step": 1530
+    },
+    {
+      "epoch": 1.6757344940152339,
+      "grad_norm": 0.02269565500319004,
+      "learning_rate": 8.697180342231849e-05,
+      "loss": 0.07186954021453858,
+      "step": 1540
+    },
+    {
+      "epoch": 1.6866158868335148,
+      "grad_norm": 0.028245631605386734,
+      "learning_rate": 8.580043550551393e-05,
+      "loss": 0.06877661943435669,
+      "step": 1550
+    },
+    {
+      "epoch": 1.6974972796517953,
+      "grad_norm": 0.02352772280573845,
+      "learning_rate": 8.463105270796276e-05,
+      "loss": 0.07183843851089478,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7083786724700762,
+      "grad_norm": 0.023996759206056595,
+      "learning_rate": 8.346381851103842e-05,
+      "loss": 0.07082110047340393,
+      "step": 1570
+    },
+    {
+      "epoch": 1.719260065288357,
+      "grad_norm": 0.025422753766179085,
+      "learning_rate": 8.229889609573699e-05,
+      "loss": 0.06895624399185181,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7301414581066377,
+      "grad_norm": 0.02499948814511299,
+      "learning_rate": 8.113644831986416e-05,
+      "loss": 0.0707672655582428,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7410228509249184,
+      "grad_norm": 0.030967354774475098,
+      "learning_rate": 7.997663769526753e-05,
+      "loss": 0.07331972122192383,
+      "step": 1600
+    },
+    {
+      "epoch": 1.7410228509249184,
+      "eval_loss": 0.07515527307987213,
+      "eval_runtime": 5.3812,
+      "eval_samples_per_second": 14.867,
+      "eval_steps_per_second": 1.858,
+      "step": 1600
+    },
+    {
+      "epoch": 1.751904243743199,
+      "grad_norm": 0.026022454723715782,
+      "learning_rate": 7.881962636511715e-05,
+      "loss": 0.07261155843734741,
+      "step": 1610
+    },
+    {
+      "epoch": 1.76278563656148,
+      "grad_norm": 0.03104907087981701,
+      "learning_rate": 7.766557608123783e-05,
+      "loss": 0.07076815962791443,
+      "step": 1620
+    },
+    {
+      "epoch": 1.7736670293797605,
+      "grad_norm": 0.026025639846920967,
+      "learning_rate": 7.651464818149587e-05,
+      "loss": 0.0643875777721405,
+      "step": 1630
+    },
+    {
+      "epoch": 1.7845484221980414,
+      "grad_norm": 0.022654540836811066,
+      "learning_rate": 7.536700356724379e-05,
+      "loss": 0.07072726488113404,
+      "step": 1640
+    },
+    {
+      "epoch": 1.795429815016322,
+      "grad_norm": 0.02240171656012535,
+      "learning_rate": 7.422280268082629e-05,
+      "loss": 0.07485873699188232,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8063112078346029,
+      "grad_norm": 0.02509446069598198,
+      "learning_rate": 7.308220548314981e-05,
+      "loss": 0.06636062860488892,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8171926006528836,
+      "grad_norm": 0.023968016728758812,
+      "learning_rate": 7.194537143132023e-05,
+      "loss": 0.06702766418457032,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8280739934711643,
+      "grad_norm": 0.027016611769795418,
+      "learning_rate": 7.081245945635013e-05,
+      "loss": 0.07087734937667847,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8389553862894452,
+      "grad_norm": 0.02593705803155899,
+      "learning_rate": 6.968362794094027e-05,
+      "loss": 0.06891222596168518,
+      "step": 1690
+    },
+    {
+      "epoch": 1.8498367791077257,
+      "grad_norm": 0.02612975798547268,
+      "learning_rate": 6.855903469733738e-05,
+      "loss": 0.06720860004425049,
+      "step": 1700
+    },
+    {
+      "epoch": 1.8607181719260066,
+      "grad_norm": 0.025018637999892235,
+      "learning_rate": 6.743883694527169e-05,
+      "loss": 0.06776891946792603,
+      "step": 1710
+    },
+    {
+      "epoch": 1.8715995647442871,
+      "grad_norm": 0.02438053861260414,
+      "learning_rate": 6.632319128997759e-05,
+      "loss": 0.07007836103439331,
+      "step": 1720
+    },
+    {
+      "epoch": 1.882480957562568,
+      "grad_norm": 0.02692045457661152,
+      "learning_rate": 6.52122537002998e-05,
+      "loss": 0.07095590829849244,
+      "step": 1730
+    },
+    {
+      "epoch": 1.8933623503808488,
+      "grad_norm": 0.031177183613181114,
+      "learning_rate": 6.410617948688892e-05,
+      "loss": 0.06855381727218628,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9042437431991295,
+      "grad_norm": 0.022553348913788795,
+      "learning_rate": 6.300512328048845e-05,
+      "loss": 0.07089903354644775,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9151251360174102,
+      "grad_norm": 0.024013947695493698,
+      "learning_rate": 6.190923901031757e-05,
+      "loss": 0.0708732008934021,
+      "step": 1760
+    },
+    {
+      "epoch": 1.926006528835691,
+      "grad_norm": 0.022148922085762024,
+      "learning_rate": 6.081867988255135e-05,
+      "loss": 0.07153573036193847,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9368879216539718,
+      "grad_norm": 0.047587718814611435,
+      "learning_rate": 5.973359835890246e-05,
+      "loss": 0.0697596788406372,
+      "step": 1780
+    },
+    {
+      "epoch": 1.9477693144722523,
+      "grad_norm": 0.02610265091061592,
+      "learning_rate": 5.865414613530669e-05,
+      "loss": 0.06653320789337158,
+      "step": 1790
+    },
+    {
+      "epoch": 1.9586507072905333,
+      "grad_norm": 0.02527604252099991,
+      "learning_rate": 5.758047412071568e-05,
+      "loss": 0.07076951861381531,
+      "step": 1800
+    },
+    {
+      "epoch": 1.9586507072905333,
+      "eval_loss": 0.07202806323766708,
+      "eval_runtime": 5.4058,
+      "eval_samples_per_second": 14.799,
+      "eval_steps_per_second": 1.85,
+      "step": 1800
+    },
+    {
+      "epoch": 1.969532100108814,
+      "grad_norm": 0.024923427030444145,
+      "learning_rate": 5.651273241599976e-05,
+      "loss": 0.06910277605056762,
+      "step": 1810
+    },
+    {
+      "epoch": 1.9804134929270947,
+      "grad_norm": 0.023715443909168243,
+      "learning_rate": 5.5451070292963434e-05,
+      "loss": 0.06721980571746826,
+      "step": 1820
+    },
+    {
+      "epoch": 1.9912948857453754,
+      "grad_norm": 0.019582392647862434,
+      "learning_rate": 5.439563617347711e-05,
+      "loss": 0.06764072179794312,
+      "step": 1830
+    },
+    {
+      "epoch": 2.002176278563656,
+      "grad_norm": 0.027950329706072807,
+      "learning_rate": 5.3346577608727463e-05,
+      "loss": 0.07124176025390624,
+      "step": 1840
+    },
+    {
+      "epoch": 2.013057671381937,
+      "grad_norm": 0.036231789737939835,
+      "learning_rate": 5.230404125858952e-05,
+      "loss": 0.05957789421081543,
+      "step": 1850
+    },
+    {
+      "epoch": 2.0239390642002175,
+      "grad_norm": 0.026198675855994225,
+      "learning_rate": 5.1268172871123535e-05,
+      "loss": 0.055424296855926515,
+      "step": 1860
+    },
+    {
+      "epoch": 2.0348204570184985,
+      "grad_norm": 0.025477459654211998,
+      "learning_rate": 5.023911726219902e-05,
+      "loss": 0.06114422678947449,
+      "step": 1870
+    },
+    {
+      "epoch": 2.045701849836779,
+      "grad_norm": 0.03552868962287903,
+      "learning_rate": 4.921701829524956e-05,
+      "loss": 0.058832204341888426,
+      "step": 1880
+    },
+    {
+      "epoch": 2.05658324265506,
+      "grad_norm": 0.03449210524559021,
+      "learning_rate": 4.820201886116022e-05,
+      "loss": 0.057713437080383304,
+      "step": 1890
+    },
+    {
+      "epoch": 2.067464635473341,
+      "grad_norm": 0.033065732568502426,
+      "learning_rate": 4.719426085829123e-05,
+      "loss": 0.05968807935714722,
+      "step": 1900
+    },
+    {
+      "epoch": 2.0783460282916213,
+      "grad_norm": 0.03002762608230114,
+      "learning_rate": 4.619388517264056e-05,
+      "loss": 0.060094785690307614,
+      "step": 1910
+    },
+    {
+      "epoch": 2.0892274211099022,
+      "grad_norm": 0.034517038613557816,
+      "learning_rate": 4.5201031658147766e-05,
+      "loss": 0.05601954460144043,
+      "step": 1920
+    },
+    {
+      "epoch": 2.1001088139281827,
+      "grad_norm": 0.03293369337916374,
+      "learning_rate": 4.4215839117142134e-05,
+      "loss": 0.06034954190254212,
+      "step": 1930
+    },
+    {
+      "epoch": 2.1109902067464636,
+      "grad_norm": 0.032126300036907196,
+      "learning_rate": 4.3238445280938154e-05,
+      "loss": 0.05510505437850952,
+      "step": 1940
+    },
+    {
+      "epoch": 2.121871599564744,
+      "grad_norm": 0.027177443727850914,
+      "learning_rate": 4.2268986790580125e-05,
+      "loss": 0.060755813121795656,
+      "step": 1950
+    },
+    {
+      "epoch": 2.132752992383025,
+      "grad_norm": 0.030124740675091743,
+      "learning_rate": 4.130759917773985e-05,
+      "loss": 0.06080647706985474,
+      "step": 1960
+    },
+    {
+      "epoch": 2.143634385201306,
+      "grad_norm": 0.03235238790512085,
+      "learning_rate": 4.0354416845769e-05,
+      "loss": 0.06485359072685241,
+      "step": 1970
+    },
+    {
+      "epoch": 2.1545157780195865,
+      "grad_norm": 0.033107586205005646,
+      "learning_rate": 3.940957305090935e-05,
+      "loss": 0.059920203685760495,
+      "step": 1980
+    },
+    {
+      "epoch": 2.1653971708378674,
+      "grad_norm": 0.02939060889184475,
+      "learning_rate": 3.8473199883663334e-05,
+      "loss": 0.0490723729133606,
+      "step": 1990
+    },
+    {
+      "epoch": 2.176278563656148,
+      "grad_norm": 0.03297962248325348,
+      "learning_rate": 3.754542825032766e-05,
+      "loss": 0.053455913066864015,
+      "step": 2000
+    },
+    {
+      "epoch": 2.176278563656148,
+      "eval_loss": 0.07033731788396835,
+      "eval_runtime": 5.4006,
+      "eval_samples_per_second": 14.813,
+      "eval_steps_per_second": 1.852,
+      "step": 2000
+    },
+    {
+      "epoch": 2.187159956474429,
+      "grad_norm": 0.026707574725151062,
+      "learning_rate": 3.6626387854692365e-05,
+      "loss": 0.061966001987457275,
+      "step": 2010
+    },
+    {
+      "epoch": 2.1980413492927093,
+      "grad_norm": 0.03455955162644386,
+      "learning_rate": 3.571620717990811e-05,
+      "loss": 0.05781112909317017,
+      "step": 2020
+    },
+    {
+      "epoch": 2.2089227421109903,
+      "grad_norm": 0.02576427161693573,
+      "learning_rate": 3.4815013470524004e-05,
+      "loss": 0.060006117820739745,
+      "step": 2030
+    },
+    {
+      "epoch": 2.2198041349292708,
+      "grad_norm": 0.03262777253985405,
+      "learning_rate": 3.392293271469866e-05,
+      "loss": 0.05959203839302063,
+      "step": 2040
+    },
+    {
+      "epoch": 2.2306855277475517,
+      "grad_norm": 0.031093895435333252,
+      "learning_rate": 3.304008962658677e-05,
+      "loss": 0.057662928104400636,
+      "step": 2050
+    },
+    {
+      "epoch": 2.2415669205658326,
+      "grad_norm": 0.030908960849046707,
+      "learning_rate": 3.216660762890413e-05,
+      "loss": 0.05794321894645691,
+      "step": 2060
+    },
+    {
+      "epoch": 2.252448313384113,
+      "grad_norm": 0.03053661808371544,
+      "learning_rate": 3.130260883567272e-05,
+      "loss": 0.05402252674102783,
+      "step": 2070
+    },
+    {
+      "epoch": 2.263329706202394,
+      "grad_norm": 0.03832336515188217,
+      "learning_rate": 3.0448214035149102e-05,
+      "loss": 0.06356992721557617,
+      "step": 2080
+    },
+    {
+      "epoch": 2.2742110990206745,
+      "grad_norm": 0.035567495971918106,
+      "learning_rate": 2.960354267293809e-05,
+      "loss": 0.057219624519348145,
+      "step": 2090
+    },
+    {
+      "epoch": 2.2850924918389555,
+      "grad_norm": 0.03574289754033089,
+      "learning_rate": 2.8768712835293922e-05,
+      "loss": 0.0654039740562439,
+      "step": 2100
+    },
+    {
+      "epoch": 2.295973884657236,
+      "grad_norm": 0.03358747810125351,
+      "learning_rate": 2.7943841232611812e-05,
+      "loss": 0.05935961604118347,
+      "step": 2110
+    },
+    {
+      "epoch": 2.306855277475517,
+      "grad_norm": 0.036444805562496185,
+      "learning_rate": 2.712904318311167e-05,
+      "loss": 0.059474557638168335,
+      "step": 2120
+    },
+    {
+      "epoch": 2.3177366702937974,
+      "grad_norm": 0.04165448620915413,
+      "learning_rate": 2.6324432596716276e-05,
+      "loss": 0.06448624134063721,
+      "step": 2130
+    },
+    {
+      "epoch": 2.3286180631120783,
+      "grad_norm": 0.03665981814265251,
+      "learning_rate": 2.5530121959126718e-05,
+      "loss": 0.055650746822357176,
+      "step": 2140
+    },
+    {
+      "epoch": 2.3394994559303592,
+      "grad_norm": 0.02873632311820984,
+      "learning_rate": 2.47462223160965e-05,
+      "loss": 0.05966170430183411,
+      "step": 2150
+    },
+    {
+      "epoch": 2.3503808487486397,
+      "grad_norm": 0.03279171139001846,
+      "learning_rate": 2.397284325790744e-05,
+      "loss": 0.06053815484046936,
+      "step": 2160
+    },
+    {
+      "epoch": 2.3612622415669207,
+      "grad_norm": 0.03719950094819069,
+      "learning_rate": 2.321009290404864e-05,
+      "loss": 0.0560091495513916,
+      "step": 2170
+    },
+    {
+      "epoch": 2.372143634385201,
+      "grad_norm": 0.03888888284564018,
+      "learning_rate": 2.2458077888101226e-05,
+      "loss": 0.058753132820129395,
+      "step": 2180
+    },
+    {
+      "epoch": 2.383025027203482,
+      "grad_norm": 0.0321931354701519,
+      "learning_rate": 2.1716903342830995e-05,
+      "loss": 0.05801019072532654,
+      "step": 2190
+    },
+    {
+      "epoch": 2.3939064200217626,
+      "grad_norm": 0.038627929985523224,
+      "learning_rate": 2.0986672885490378e-05,
+      "loss": 0.06117302179336548,
+      "step": 2200
+    },
+    {
+      "epoch": 2.3939064200217626,
+      "eval_loss": 0.0698942095041275,
+      "eval_runtime": 5.4996,
+      "eval_samples_per_second": 14.546,
+      "eval_steps_per_second": 1.818,
+      "step": 2200
+    },
+    {
+      "epoch": 2.4047878128400435,
+      "grad_norm": 0.03631049767136574,
+      "learning_rate": 2.0267488603332907e-05,
+      "loss": 0.05466359853744507,
+      "step": 2210
+    },
+    {
+      "epoch": 2.4156692056583244,
+      "grad_norm": 0.03237799555063248,
+      "learning_rate": 1.9559451039341147e-05,
+      "loss": 0.05731322765350342,
+      "step": 2220
+    },
+    {
+      "epoch": 2.426550598476605,
+      "grad_norm": 0.0349552221596241,
+      "learning_rate": 1.886265917817064e-05,
+      "loss": 0.06014159917831421,
+      "step": 2230
+    },
+    {
+      "epoch": 2.437431991294886,
+      "grad_norm": 0.0360136479139328,
+      "learning_rate": 1.8177210432311732e-05,
+      "loss": 0.05688718557357788,
+      "step": 2240
+    },
+    {
+      "epoch": 2.4483133841131663,
+      "grad_norm": 0.03673616796731949,
+      "learning_rate": 1.750320062847116e-05,
+      "loss": 0.058925163745880124,
+      "step": 2250
+    },
+    {
+      "epoch": 2.4591947769314473,
+      "grad_norm": 0.050142526626586914,
+      "learning_rate": 1.6840723994175356e-05,
+      "loss": 0.05712746381759644,
+      "step": 2260
+    },
+    {
+      "epoch": 2.4700761697497278,
+      "grad_norm": 0.03750143572688103,
+      "learning_rate": 1.6189873144597267e-05,
+      "loss": 0.056266939640045165,
+      "step": 2270
+    },
+    {
+      "epoch": 2.4809575625680087,
+      "grad_norm": 0.03403429314494133,
+      "learning_rate": 1.5550739069608676e-05,
+      "loss": 0.05650065541267395,
+      "step": 2280
+    },
+    {
+      "epoch": 2.4918389553862896,
+      "grad_norm": 0.030664529651403427,
+      "learning_rate": 1.492341112105966e-05,
+      "loss": 0.05500165820121765,
+      "step": 2290
+    },
+    {
+      "epoch": 2.50272034820457,
+      "grad_norm": 0.03341777250170708,
+      "learning_rate": 1.4307977000287064e-05,
+      "loss": 0.05969707369804382,
+      "step": 2300
+    },
+    {
+      "epoch": 2.513601741022851,
+      "grad_norm": 0.03309670463204384,
+      "learning_rate": 1.3704522745853743e-05,
+      "loss": 0.05166618824005127,
+      "step": 2310
+    },
+    {
+      "epoch": 2.5244831338411315,
+      "grad_norm": 0.03246740996837616,
+      "learning_rate": 1.311313272152025e-05,
+      "loss": 0.0567930817604065,
+      "step": 2320
+    },
+    {
+      "epoch": 2.5353645266594125,
+      "grad_norm": 0.036715373396873474,
+      "learning_rate": 1.2533889604450677e-05,
+      "loss": 0.05777972936630249,
+      "step": 2330
+    },
+    {
+      "epoch": 2.546245919477693,
+      "grad_norm": 0.03496614098548889,
+      "learning_rate": 1.196687437365427e-05,
+      "loss": 0.054594647884368894,
+      "step": 2340
+    },
+    {
+      "epoch": 2.557127312295974,
+      "grad_norm": 0.03273237124085426,
+      "learning_rate": 1.141216629866434e-05,
+      "loss": 0.05491857528686524,
+      "step": 2350
+    },
+    {
+      "epoch": 2.568008705114255,
+      "grad_norm": 0.03323022276163101,
+      "learning_rate": 1.0869842928456441e-05,
+      "loss": 0.05289023518562317,
+      "step": 2360
+    },
+    {
+      "epoch": 2.5788900979325353,
+      "grad_norm": 0.042828429490327835,
+      "learning_rate": 1.033998008060687e-05,
+      "loss": 0.05806971788406372,
+      "step": 2370
+    },
+    {
+      "epoch": 2.5897714907508163,
+      "grad_norm": 0.03340109810233116,
+      "learning_rate": 9.822651830693152e-06,
+      "loss": 0.05867668390274048,
+      "step": 2380
+    },
+    {
+      "epoch": 2.6006528835690967,
+      "grad_norm": 0.03182734549045563,
+      "learning_rate": 9.317930501938289e-06,
+      "loss": 0.05183411836624145,
+      "step": 2390
+    },
+    {
+      "epoch": 2.6115342763873777,
+      "grad_norm": 0.03279723972082138,
+      "learning_rate": 8.82588665509978e-06,
+      "loss": 0.06119837760925293,
+      "step": 2400
+    },
+    {
+      "epoch": 2.6115342763873777,
+      "eval_loss": 0.06727322936058044,
+      "eval_runtime": 5.4288,
+      "eval_samples_per_second": 14.736,
+      "eval_steps_per_second": 1.842,
+      "step": 2400
+    },
+    {
+      "epoch": 2.622415669205658,
+      "grad_norm": 0.03709235042333603,
+      "learning_rate": 8.346589078605215e-06,
+      "loss": 0.05637891292572021,
+      "step": 2410
+    },
+    {
+      "epoch": 2.633297062023939,
+      "grad_norm": 0.030506664887070656,
+      "learning_rate": 7.880104778935505e-06,
+      "loss": 0.05910442471504211,
+      "step": 2420
+    },
+    {
+      "epoch": 2.64417845484222,
+      "grad_norm": 0.03575789928436279,
+      "learning_rate": 7.426498971257278e-06,
+      "loss": 0.054590433835983276,
+      "step": 2430
+    },
+    {
+      "epoch": 2.6550598476605005,
+      "grad_norm": 0.035656314343214035,
+      "learning_rate": 6.985835070305768e-06,
+      "loss": 0.056293058395385745,
+      "step": 2440
+    },
+    {
+      "epoch": 2.6659412404787814,
+      "grad_norm": 0.03722505271434784,
+      "learning_rate": 6.5581746815192135e-06,
+      "loss": 0.058118653297424314,
+      "step": 2450
+    },
+    {
+      "epoch": 2.676822633297062,
+      "grad_norm": 0.035609811544418335,
+      "learning_rate": 6.143577592426453e-06,
+      "loss": 0.05446176528930664,
+      "step": 2460
+    },
+    {
+      "epoch": 2.687704026115343,
+      "grad_norm": 0.035058531910181046,
+      "learning_rate": 5.742101764288521e-06,
+      "loss": 0.05432036519050598,
+      "step": 2470
+    },
+    {
+      "epoch": 2.6985854189336234,
+      "grad_norm": 0.036867234855890274,
+      "learning_rate": 5.353803323995521e-06,
+      "loss": 0.05071129202842713,
+      "step": 2480
+    },
+    {
+      "epoch": 2.7094668117519043,
+      "grad_norm": 0.030480151996016502,
+      "learning_rate": 4.9787365562200985e-06,
+      "loss": 0.04942301511764526,
+      "step": 2490
+    },
+    {
+      "epoch": 2.7203482045701852,
+      "grad_norm": 0.0435580313205719,
+      "learning_rate": 4.616953895828269e-06,
+      "loss": 0.05862914323806763,
+      "step": 2500
+    },
+    {
+      "epoch": 2.7312295973884657,
+      "grad_norm": 0.03596170246601105,
+      "learning_rate": 4.268505920549016e-06,
+      "loss": 0.054695284366607665,
+      "step": 2510
+    },
+    {
+      "epoch": 2.742110990206746,
+      "grad_norm": 0.034583065658807755,
+      "learning_rate": 3.933441343903444e-06,
+      "loss": 0.04879617691040039,
+      "step": 2520
+    },
+    {
+      "epoch": 2.752992383025027,
+      "grad_norm": 0.03540559113025665,
+      "learning_rate": 3.61180700839453e-06,
+      "loss": 0.05482098460197449,
+      "step": 2530
+    },
+    {
+      "epoch": 2.763873775843308,
+      "grad_norm": 0.0387350358068943,
+      "learning_rate": 3.3036478789585044e-06,
+      "loss": 0.05537814497947693,
+      "step": 2540
+    },
+    {
+      "epoch": 2.7747551686615886,
+      "grad_norm": 0.03608511760830879,
+      "learning_rate": 3.0090070366786792e-06,
+      "loss": 0.05486249327659607,
+      "step": 2550
+    },
+    {
+      "epoch": 2.7856365614798695,
+      "grad_norm": 0.03293207287788391,
+      "learning_rate": 2.7279256727626546e-06,
+      "loss": 0.054778552055358885,
+      "step": 2560
+    },
+    {
+      "epoch": 2.7965179542981504,
+      "grad_norm": 0.03599146753549576,
+      "learning_rate": 2.4604430827837143e-06,
+      "loss": 0.05409228205680847,
+      "step": 2570
+    },
+    {
+      "epoch": 2.807399347116431,
+      "grad_norm": 0.03957504406571388,
+      "learning_rate": 2.2065966611872657e-06,
+      "loss": 0.06026080846786499,
+      "step": 2580
+    },
+    {
+      "epoch": 2.8182807399347114,
+      "grad_norm": 0.033959709107875824,
+      "learning_rate": 1.9664218960630533e-06,
+      "loss": 0.05407561063766479,
+      "step": 2590
+    },
+    {
+      "epoch": 2.8291621327529923,
+      "grad_norm": 0.04043720290064812,
+      "learning_rate": 1.7399523641838278e-06,
+      "loss": 0.0522362232208252,
+      "step": 2600
+    },
+    {
+      "epoch": 2.8291621327529923,
+      "eval_loss": 0.06709529459476471,
+      "eval_runtime": 5.6005,
+      "eval_samples_per_second": 14.284,
+      "eval_steps_per_second": 1.786,
+      "step": 2600
+    },
+    {
+      "epoch": 2.8400435255712733,
+      "grad_norm": 0.034862909466028214,
+      "learning_rate": 1.5272197263113131e-06,
+      "loss": 0.050741416215896604,
+      "step": 2610
+    },
+    {
+      "epoch": 2.8509249183895538,
+      "grad_norm": 0.03554687649011612,
+      "learning_rate": 1.3282537227699787e-06,
+      "loss": 0.054420065879821775,
+      "step": 2620
+    },
+    {
+      "epoch": 2.8618063112078347,
+      "grad_norm": 0.0366034209728241,
+      "learning_rate": 1.143082169289289e-06,
+      "loss": 0.05170666575431824,
+      "step": 2630
+    },
+    {
+      "epoch": 2.8726877040261156,
+      "grad_norm": 0.03735072910785675,
+      "learning_rate": 9.717309531150132e-07,
+      "loss": 0.05548094511032105,
+      "step": 2640
+    },
+    {
+      "epoch": 2.883569096844396,
+      "grad_norm": 0.034383635967969894,
+      "learning_rate": 8.142240293901649e-07,
+      "loss": 0.05571337938308716,
+      "step": 2650
+    },
+    {
+      "epoch": 2.8944504896626766,
+      "grad_norm": 0.03764410689473152,
+      "learning_rate": 6.705834178060589e-07,
+      "loss": 0.05904418230056763,
+      "step": 2660
+    },
+    {
+      "epoch": 2.9053318824809575,
+      "grad_norm": 0.03647703677415848,
+      "learning_rate": 5.408291995238956e-07,
+      "loss": 0.0517579436302185,
+      "step": 2670
+    },
+    {
+      "epoch": 2.9162132752992385,
+      "grad_norm": 0.033308230340480804,
+      "learning_rate": 4.249795143673962e-07,
+      "loss": 0.05420621633529663,
+      "step": 2680
+    },
+    {
+      "epoch": 2.927094668117519,
+      "grad_norm": 0.03559701517224312,
+      "learning_rate": 3.230505582868637e-07,
+      "loss": 0.060290944576263425,
+      "step": 2690
+    },
+    {
+      "epoch": 2.9379760609358,
+      "grad_norm": 0.03408787399530411,
+      "learning_rate": 2.3505658109487238e-07,
+      "loss": 0.05462511777877808,
+      "step": 2700
+    },
+    {
+      "epoch": 2.9488574537540804,
+      "grad_norm": 0.036027781665325165,
+      "learning_rate": 1.6100988447423915e-07,
+      "loss": 0.06044437885284424,
+      "step": 2710
+    },
+    {
+      "epoch": 2.9597388465723613,
+      "grad_norm": 0.03578963875770569,
+      "learning_rate": 1.0092082025815464e-07,
+      "loss": 0.0504527747631073,
+      "step": 2720
+    },
+    {
+      "epoch": 2.970620239390642,
+      "grad_norm": 0.03743607550859451,
+      "learning_rate": 5.479778898302979e-08,
+      "loss": 0.053725385665893556,
+      "step": 2730
+    },
+    {
+      "epoch": 2.9815016322089227,
+      "grad_norm": 0.03584102541208267,
+      "learning_rate": 2.2647238714024188e-08,
+      "loss": 0.05662902593612671,
+      "step": 2740
+    },
+    {
+      "epoch": 2.9923830250272037,
+      "grad_norm": 0.029956763610243797,
+      "learning_rate": 4.473664143667122e-09,
+      "loss": 0.0520307719707489,
+      "step": 2750
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2757,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.480981168874848e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

qwen3_8b_sft_rtl/checkpoint-2757/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63e1337140001fd786f5bfde65ea532ba76972f9e8e0826ca7b80b663dbc5302
+size 5176

qwen3_8b_sft_rtl/eval_results.json

@@ -0,0 +1,7 @@
+{
+    "epoch": 3.0,
+    "eval_loss": 0.06705856323242188,
+    "eval_runtime": 5.4283,
+    "eval_samples_per_second": 14.738,
+    "eval_steps_per_second": 1.842
+}

qwen3_8b_sft_rtl/tokenizer.json

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:be75606093db2094d7cd20f3c2f385c212750648bd6ea4fb2bf507a6a4c55506
+size 11422650

qwen3_8b_sft_rtl/tokenizer_config.json

@@ -0,0 +1,30 @@
+{
+  "add_prefix_space": false,
+  "backend": "tokenizers",
+  "bos_token": null,
+  "clean_up_tokenization_spaces": false,
+  "eos_token": "<|im_end|>",
+  "errors": "replace",
+  "extra_special_tokens": [
+    "<|im_start|>",
+    "<|im_end|>",
+    "<|object_ref_start|>",
+    "<|object_ref_end|>",
+    "<|box_start|>",
+    "<|box_end|>",
+    "<|quad_start|>",
+    "<|quad_end|>",
+    "<|vision_start|>",
+    "<|vision_end|>",
+    "<|vision_pad|>",
+    "<|image_pad|>",
+    "<|video_pad|>"
+  ],
+  "is_local": true,
+  "model_max_length": 131072,
+  "pad_token": "<|endoftext|>",
+  "padding_side": "right",
+  "split_special_tokens": false,
+  "tokenizer_class": "Qwen2Tokenizer",
+  "unk_token": null
+}

qwen3_8b_sft_rtl/train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 3.0,
+    "total_flos": 4.480981168874848e+18,
+    "train_loss": 0.07706532943797881,
+    "train_runtime": 10631.5322,
+    "train_samples_per_second": 4.149,
+    "train_steps_per_second": 0.259
+}

qwen3_8b_sft_rtl/trainer_log.jsonl

@@ -0,0 +1,289 @@
+{"current_steps": 10, "total_steps": 2757, "loss": 0.741086769104004, "lr": 1.8e-05, "epoch": 0.01088139281828074, "percentage": 0.36, "elapsed_time": "0:00:40", "remaining_time": "3:07:36"}
+{"current_steps": 20, "total_steps": 2757, "loss": 0.3913477897644043, "lr": 3.8e-05, "epoch": 0.02176278563656148, "percentage": 0.73, "elapsed_time": "0:01:19", "remaining_time": "3:01:59"}
+{"current_steps": 30, "total_steps": 2757, "loss": 0.22936077117919923, "lr": 5.8e-05, "epoch": 0.03264417845484222, "percentage": 1.09, "elapsed_time": "0:01:57", "remaining_time": "2:58:20"}
+{"current_steps": 40, "total_steps": 2757, "loss": 0.17587745189666748, "lr": 7.800000000000001e-05, "epoch": 0.04352557127312296, "percentage": 1.45, "elapsed_time": "0:02:35", "remaining_time": "2:56:07"}
+{"current_steps": 50, "total_steps": 2757, "loss": 0.14949668645858766, "lr": 9.8e-05, "epoch": 0.0544069640914037, "percentage": 1.81, "elapsed_time": "0:03:13", "remaining_time": "2:54:35"}
+{"current_steps": 60, "total_steps": 2757, "loss": 0.13353660106658935, "lr": 0.000118, "epoch": 0.06528835690968444, "percentage": 2.18, "elapsed_time": "0:03:51", "remaining_time": "2:53:28"}
+{"current_steps": 70, "total_steps": 2757, "loss": 0.120628023147583, "lr": 0.000138, "epoch": 0.07616974972796518, "percentage": 2.54, "elapsed_time": "0:04:29", "remaining_time": "2:52:08"}
+{"current_steps": 80, "total_steps": 2757, "loss": 0.11396801471710205, "lr": 0.00015800000000000002, "epoch": 0.08705114254624592, "percentage": 2.9, "elapsed_time": "0:05:17", "remaining_time": "2:57:01"}
+{"current_steps": 90, "total_steps": 2757, "loss": 0.1074080228805542, "lr": 0.00017800000000000002, "epoch": 0.09793253536452666, "percentage": 3.26, "elapsed_time": "0:05:56", "remaining_time": "2:56:01"}
+{"current_steps": 100, "total_steps": 2757, "loss": 0.10922808647155761, "lr": 0.00019800000000000002, "epoch": 0.1088139281828074, "percentage": 3.63, "elapsed_time": "0:06:34", "remaining_time": "2:54:34"}
+{"current_steps": 110, "total_steps": 2757, "loss": 0.10778391361236572, "lr": 0.00019999433803003197, "epoch": 0.11969532100108814, "percentage": 3.99, "elapsed_time": "0:07:11", "remaining_time": "2:53:06"}
+{"current_steps": 120, "total_steps": 2757, "loss": 0.10001739263534545, "lr": 0.00019997476661131597, "epoch": 0.1305767138193689, "percentage": 4.35, "elapsed_time": "0:07:49", "remaining_time": "2:51:56"}
+{"current_steps": 130, "total_steps": 2757, "loss": 0.10446863174438477, "lr": 0.0001999412185784487, "epoch": 0.14145810663764963, "percentage": 4.72, "elapsed_time": "0:08:27", "remaining_time": "2:50:48"}
+{"current_steps": 140, "total_steps": 2757, "loss": 0.10438344478607178, "lr": 0.0001998936986214928, "epoch": 0.15233949945593037, "percentage": 5.08, "elapsed_time": "0:09:04", "remaining_time": "2:49:44"}
+{"current_steps": 150, "total_steps": 2757, "loss": 0.09951521158218384, "lr": 0.00019983221338380563, "epoch": 0.1632208922742111, "percentage": 5.44, "elapsed_time": "0:09:42", "remaining_time": "2:48:38"}
+{"current_steps": 160, "total_steps": 2757, "loss": 0.09911214709281921, "lr": 0.0001997567714611106, "epoch": 0.17410228509249184, "percentage": 5.8, "elapsed_time": "0:10:19", "remaining_time": "2:47:37"}
+{"current_steps": 170, "total_steps": 2757, "loss": 0.09334277510643005, "lr": 0.0001996673834002955, "epoch": 0.18498367791077258, "percentage": 6.17, "elapsed_time": "0:10:59", "remaining_time": "2:47:14"}
+{"current_steps": 180, "total_steps": 2757, "loss": 0.09431872367858887, "lr": 0.00019956406169793805, "epoch": 0.19586507072905332, "percentage": 6.53, "elapsed_time": "0:11:37", "remaining_time": "2:46:20"}
+{"current_steps": 190, "total_steps": 2757, "loss": 0.10089273452758789, "lr": 0.00019944682079855867, "epoch": 0.20674646354733406, "percentage": 6.89, "elapsed_time": "0:12:15", "remaining_time": "2:45:32"}
+{"current_steps": 200, "total_steps": 2757, "loss": 0.0949325442314148, "lr": 0.0001993156770926014, "epoch": 0.2176278563656148, "percentage": 7.25, "elapsed_time": "0:12:52", "remaining_time": "2:44:38"}
+{"current_steps": 200, "total_steps": 2757, "eval_loss": 0.09785260260105133, "epoch": 0.2176278563656148, "percentage": 7.25, "elapsed_time": "0:12:58", "remaining_time": "2:45:48"}
+{"current_steps": 210, "total_steps": 2757, "loss": 0.09331356287002564, "lr": 0.00019917064891414232, "epoch": 0.22850924918389554, "percentage": 7.62, "elapsed_time": "0:13:44", "remaining_time": "2:46:41"}
+{"current_steps": 220, "total_steps": 2757, "loss": 0.09654195308685302, "lr": 0.00019901175653832645, "epoch": 0.23939064200217627, "percentage": 7.98, "elapsed_time": "0:14:22", "remaining_time": "2:45:45"}
+{"current_steps": 230, "total_steps": 2757, "loss": 0.09114633202552795, "lr": 0.00019883902217853326, "epoch": 0.250272034820457, "percentage": 8.34, "elapsed_time": "0:15:00", "remaining_time": "2:44:55"}
+{"current_steps": 240, "total_steps": 2757, "loss": 0.09384369254112243, "lr": 0.0001986524699832712, "epoch": 0.2611534276387378, "percentage": 8.71, "elapsed_time": "0:15:38", "remaining_time": "2:44:04"}
+{"current_steps": 250, "total_steps": 2757, "loss": 0.08879860639572143, "lr": 0.00019845212603280174, "epoch": 0.2720348204570185, "percentage": 9.07, "elapsed_time": "0:16:16", "remaining_time": "2:43:09"}
+{"current_steps": 260, "total_steps": 2757, "loss": 0.09065196514129639, "lr": 0.00019823801833549327, "epoch": 0.28291621327529926, "percentage": 9.43, "elapsed_time": "0:16:53", "remaining_time": "2:42:15"}
+{"current_steps": 270, "total_steps": 2757, "loss": 0.09161882400512696, "lr": 0.00019801017682390552, "epoch": 0.29379760609357997, "percentage": 9.79, "elapsed_time": "0:17:30", "remaining_time": "2:41:20"}
+{"current_steps": 280, "total_steps": 2757, "loss": 0.09159514904022217, "lr": 0.0001977686333506049, "epoch": 0.30467899891186073, "percentage": 10.16, "elapsed_time": "0:18:08", "remaining_time": "2:40:32"}
+{"current_steps": 290, "total_steps": 2757, "loss": 0.0902558445930481, "lr": 0.00019751342168371154, "epoch": 0.31556039173014144, "percentage": 10.52, "elapsed_time": "0:18:46", "remaining_time": "2:39:43"}
+{"current_steps": 300, "total_steps": 2757, "loss": 0.09181964993476868, "lr": 0.00019724457750217839, "epoch": 0.3264417845484222, "percentage": 10.88, "elapsed_time": "0:19:24", "remaining_time": "2:38:53"}
+{"current_steps": 310, "total_steps": 2757, "loss": 0.08723740577697754, "lr": 0.0001969621383908033, "epoch": 0.3373231773667029, "percentage": 11.24, "elapsed_time": "0:20:01", "remaining_time": "2:38:02"}
+{"current_steps": 320, "total_steps": 2757, "loss": 0.09164115190505981, "lr": 0.00019666614383497457, "epoch": 0.3482045701849837, "percentage": 11.61, "elapsed_time": "0:20:38", "remaining_time": "2:37:13"}
+{"current_steps": 330, "total_steps": 2757, "loss": 0.08903414011001587, "lr": 0.00019635663521515087, "epoch": 0.3590859630032644, "percentage": 11.97, "elapsed_time": "0:21:16", "remaining_time": "2:36:24"}
+{"current_steps": 340, "total_steps": 2757, "loss": 0.08611454963684081, "lr": 0.00019603365580107615, "epoch": 0.36996735582154516, "percentage": 12.33, "elapsed_time": "0:22:04", "remaining_time": "2:36:58"}
+{"current_steps": 350, "total_steps": 2757, "loss": 0.08958455920219421, "lr": 0.0001956972507457306, "epoch": 0.3808487486398259, "percentage": 12.69, "elapsed_time": "0:22:42", "remaining_time": "2:36:11"}
+{"current_steps": 360, "total_steps": 2757, "loss": 0.08454820513725281, "lr": 0.000195347467079018, "epoch": 0.39173014145810664, "percentage": 13.06, "elapsed_time": "0:23:20", "remaining_time": "2:35:23"}
+{"current_steps": 370, "total_steps": 2757, "loss": 0.09067847728729247, "lr": 0.00019498435370119094, "epoch": 0.40261153427638735, "percentage": 13.42, "elapsed_time": "0:23:59", "remaining_time": "2:34:49"}
+{"current_steps": 380, "total_steps": 2757, "loss": 0.0873767077922821, "lr": 0.00019460796137601467, "epoch": 0.4134929270946681, "percentage": 13.78, "elapsed_time": "0:24:38", "remaining_time": "2:34:05"}
+{"current_steps": 390, "total_steps": 2757, "loss": 0.08544692993164063, "lr": 0.00019421834272367, "epoch": 0.4243743199129489, "percentage": 14.15, "elapsed_time": "0:25:15", "remaining_time": "2:33:18"}
+{"current_steps": 400, "total_steps": 2757, "loss": 0.08988077044487, "lr": 0.00019381555221339707, "epoch": 0.4352557127312296, "percentage": 14.51, "elapsed_time": "0:25:53", "remaining_time": "2:32:33"}
+{"current_steps": 400, "total_steps": 2757, "eval_loss": 0.09069998562335968, "epoch": 0.4352557127312296, "percentage": 14.51, "elapsed_time": "0:25:58", "remaining_time": "2:33:05"}
+{"current_steps": 410, "total_steps": 2757, "loss": 0.088669753074646, "lr": 0.00019339964615588043, "epoch": 0.44613710554951036, "percentage": 14.87, "elapsed_time": "0:26:44", "remaining_time": "2:33:05"}
+{"current_steps": 420, "total_steps": 2757, "loss": 0.09294685125350952, "lr": 0.0001929706826953767, "epoch": 0.45701849836779107, "percentage": 15.23, "elapsed_time": "0:27:22", "remaining_time": "2:32:18"}
+{"current_steps": 430, "total_steps": 2757, "loss": 0.08597595095634461, "lr": 0.00019252872180158605, "epoch": 0.46789989118607184, "percentage": 15.6, "elapsed_time": "0:28:00", "remaining_time": "2:31:32"}
+{"current_steps": 440, "total_steps": 2757, "loss": 0.08706499338150024, "lr": 0.0001920738252612682, "epoch": 0.47878128400435255, "percentage": 15.96, "elapsed_time": "0:28:39", "remaining_time": "2:30:53"}
+{"current_steps": 450, "total_steps": 2757, "loss": 0.08425920009613037, "lr": 0.0001916060566696045, "epoch": 0.4896626768226333, "percentage": 16.32, "elapsed_time": "0:29:16", "remaining_time": "2:30:06"}
+{"current_steps": 460, "total_steps": 2757, "loss": 0.08545957803726197, "lr": 0.00019112548142130757, "epoch": 0.500544069640914, "percentage": 16.68, "elapsed_time": "0:29:54", "remaining_time": "2:29:19"}
+{"current_steps": 470, "total_steps": 2757, "loss": 0.08239468932151794, "lr": 0.00019063216670147854, "epoch": 0.5114254624591947, "percentage": 17.05, "elapsed_time": "0:30:31", "remaining_time": "2:28:32"}
+{"current_steps": 480, "total_steps": 2757, "loss": 0.08288769721984864, "lr": 0.00019012618147621488, "epoch": 0.5223068552774756, "percentage": 17.41, "elapsed_time": "0:31:08", "remaining_time": "2:27:45"}
+{"current_steps": 490, "total_steps": 2757, "loss": 0.08916928768157958, "lr": 0.00018960759648296865, "epoch": 0.5331882480957563, "percentage": 17.77, "elapsed_time": "0:31:46", "remaining_time": "2:26:59"}
+{"current_steps": 500, "total_steps": 2757, "loss": 0.09119058847427368, "lr": 0.0001890764842206573, "epoch": 0.544069640914037, "percentage": 18.14, "elapsed_time": "0:32:24", "remaining_time": "2:26:16"}
+{"current_steps": 510, "total_steps": 2757, "loss": 0.08246206045150757, "lr": 0.00018853291893952833, "epoch": 0.5549510337323177, "percentage": 18.5, "elapsed_time": "0:33:02", "remaining_time": "2:25:33"}
+{"current_steps": 520, "total_steps": 2757, "loss": 0.0803530752658844, "lr": 0.0001879769766307789, "epoch": 0.5658324265505985, "percentage": 18.86, "elapsed_time": "0:33:39", "remaining_time": "2:24:49"}
+{"current_steps": 530, "total_steps": 2757, "loss": 0.08848085403442382, "lr": 0.00018740873501593214, "epoch": 0.5767138193688792, "percentage": 19.22, "elapsed_time": "0:34:17", "remaining_time": "2:24:04"}
+{"current_steps": 540, "total_steps": 2757, "loss": 0.08710720539093017, "lr": 0.0001868282735359717, "epoch": 0.5875952121871599, "percentage": 19.59, "elapsed_time": "0:34:55", "remaining_time": "2:23:21"}
+{"current_steps": 550, "total_steps": 2757, "loss": 0.07982800006866456, "lr": 0.00018623567334023563, "epoch": 0.5984766050054406, "percentage": 19.95, "elapsed_time": "0:35:32", "remaining_time": "2:22:36"}
+{"current_steps": 560, "total_steps": 2757, "loss": 0.08492777943611145, "lr": 0.0001856310172750717, "epoch": 0.6093579978237215, "percentage": 20.31, "elapsed_time": "0:36:11", "remaining_time": "2:21:57"}
+{"current_steps": 570, "total_steps": 2757, "loss": 0.08290873765945435, "lr": 0.00018501438987225536, "epoch": 0.6202393906420022, "percentage": 20.67, "elapsed_time": "0:36:49", "remaining_time": "2:21:15"}
+{"current_steps": 580, "total_steps": 2757, "loss": 0.08711419701576233, "lr": 0.00018438587733717187, "epoch": 0.6311207834602829, "percentage": 21.04, "elapsed_time": "0:37:26", "remaining_time": "2:20:32"}
+{"current_steps": 590, "total_steps": 2757, "loss": 0.08342958688735962, "lr": 0.00018374556753676506, "epoch": 0.6420021762785637, "percentage": 21.4, "elapsed_time": "0:38:04", "remaining_time": "2:19:49"}
+{"current_steps": 600, "total_steps": 2757, "loss": 0.07860854864120484, "lr": 0.00018309354998725306, "epoch": 0.6528835690968444, "percentage": 21.76, "elapsed_time": "0:38:41", "remaining_time": "2:19:05"}
+{"current_steps": 600, "total_steps": 2757, "eval_loss": 0.08754301071166992, "epoch": 0.6528835690968444, "percentage": 21.76, "elapsed_time": "0:38:46", "remaining_time": "2:19:25"}
+{"current_steps": 610, "total_steps": 2757, "loss": 0.08734351992607117, "lr": 0.00018242991584161393, "epoch": 0.6637649619151251, "percentage": 22.13, "elapsed_time": "0:39:34", "remaining_time": "2:19:16"}
+{"current_steps": 620, "total_steps": 2757, "loss": 0.08437882661819458, "lr": 0.0001817547578768424, "epoch": 0.6746463547334058, "percentage": 22.49, "elapsed_time": "0:40:11", "remaining_time": "2:18:32"}
+{"current_steps": 630, "total_steps": 2757, "loss": 0.08016583323478699, "lr": 0.00018106817048097937, "epoch": 0.6855277475516867, "percentage": 22.85, "elapsed_time": "0:40:48", "remaining_time": "2:17:47"}
+{"current_steps": 640, "total_steps": 2757, "loss": 0.076961088180542, "lr": 0.0001803702496399165, "epoch": 0.6964091403699674, "percentage": 23.21, "elapsed_time": "0:41:26", "remaining_time": "2:17:04"}
+{"current_steps": 650, "total_steps": 2757, "loss": 0.08545655012130737, "lr": 0.0001796610929239771, "epoch": 0.7072905331882481, "percentage": 23.58, "elapsed_time": "0:42:06", "remaining_time": "2:16:29"}
+{"current_steps": 660, "total_steps": 2757, "loss": 0.0809054434299469, "lr": 0.00017894079947427573, "epoch": 0.7181719260065288, "percentage": 23.94, "elapsed_time": "0:42:45", "remaining_time": "2:15:50"}
+{"current_steps": 670, "total_steps": 2757, "loss": 0.07935836911201477, "lr": 0.0001782094699888582, "epoch": 0.7290533188248096, "percentage": 24.3, "elapsed_time": "0:43:23", "remaining_time": "2:15:08"}
+{"current_steps": 680, "total_steps": 2757, "loss": 0.08474135994911194, "lr": 0.00017746720670862363, "epoch": 0.7399347116430903, "percentage": 24.66, "elapsed_time": "0:44:00", "remaining_time": "2:14:26"}
+{"current_steps": 690, "total_steps": 2757, "loss": 0.08088613748550415, "lr": 0.0001767141134030313, "epoch": 0.750816104461371, "percentage": 25.03, "elapsed_time": "0:44:38", "remaining_time": "2:13:43"}
+{"current_steps": 700, "total_steps": 2757, "loss": 0.078003191947937, "lr": 0.00017595029535559336, "epoch": 0.7616974972796517, "percentage": 25.39, "elapsed_time": "0:45:16", "remaining_time": "2:13:01"}
+{"current_steps": 710, "total_steps": 2757, "loss": 0.08291269540786743, "lr": 0.00017517585934915595, "epoch": 0.7725788900979326, "percentage": 25.75, "elapsed_time": "0:45:53", "remaining_time": "2:12:18"}
+{"current_steps": 720, "total_steps": 2757, "loss": 0.08375194072723388, "lr": 0.00017439091365097106, "epoch": 0.7834602829162133, "percentage": 26.12, "elapsed_time": "0:46:34", "remaining_time": "2:11:44"}
+{"current_steps": 730, "total_steps": 2757, "loss": 0.08069210052490235, "lr": 0.00017359556799756052, "epoch": 0.794341675734494, "percentage": 26.48, "elapsed_time": "0:47:11", "remaining_time": "2:11:02"}
+{"current_steps": 740, "total_steps": 2757, "loss": 0.08342626094818115, "lr": 0.00017278993357937455, "epoch": 0.8052230685527747, "percentage": 26.84, "elapsed_time": "0:47:49", "remaining_time": "2:10:21"}
+{"current_steps": 750, "total_steps": 2757, "loss": 0.0814591646194458, "lr": 0.0001719741230252473, "epoch": 0.8161044613710555, "percentage": 27.2, "elapsed_time": "0:48:26", "remaining_time": "2:09:38"}
+{"current_steps": 760, "total_steps": 2757, "loss": 0.08308478593826293, "lr": 0.00017114825038665123, "epoch": 0.8269858541893362, "percentage": 27.57, "elapsed_time": "0:49:05", "remaining_time": "2:09:00"}
+{"current_steps": 770, "total_steps": 2757, "loss": 0.07733856439590454, "lr": 0.0001703124311217524, "epoch": 0.8378672470076169, "percentage": 27.93, "elapsed_time": "0:49:43", "remaining_time": "2:08:19"}
+{"current_steps": 780, "total_steps": 2757, "loss": 0.0828587532043457, "lr": 0.00016946678207926938, "epoch": 0.8487486398258978, "percentage": 28.29, "elapsed_time": "0:50:20", "remaining_time": "2:07:36"}
+{"current_steps": 790, "total_steps": 2757, "loss": 0.07782068252563476, "lr": 0.0001686114214821375, "epoch": 0.8596300326441785, "percentage": 28.65, "elapsed_time": "0:50:57", "remaining_time": "2:06:53"}
+{"current_steps": 800, "total_steps": 2757, "loss": 0.08273504972457886, "lr": 0.00016774646891098134, "epoch": 0.8705114254624592, "percentage": 29.02, "elapsed_time": "0:51:36", "remaining_time": "2:06:13"}
+{"current_steps": 800, "total_steps": 2757, "eval_loss": 0.08349094539880753, "epoch": 0.8705114254624592, "percentage": 29.02, "elapsed_time": "0:51:41", "remaining_time": "2:06:27"}
+{"current_steps": 810, "total_steps": 2757, "loss": 0.08457955718040466, "lr": 0.00016687204528739694, "epoch": 0.8813928182807399, "percentage": 29.38, "elapsed_time": "0:52:28", "remaining_time": "2:06:07"}
+{"current_steps": 820, "total_steps": 2757, "loss": 0.08319224119186401, "lr": 0.00016598827285704696, "epoch": 0.8922742110990207, "percentage": 29.74, "elapsed_time": "0:53:05", "remaining_time": "2:05:25"}
+{"current_steps": 830, "total_steps": 2757, "loss": 0.08071508407592773, "lr": 0.00016509527517257042, "epoch": 0.9031556039173014, "percentage": 30.11, "elapsed_time": "0:53:43", "remaining_time": "2:04:43"}
+{"current_steps": 840, "total_steps": 2757, "loss": 0.08242483139038086, "lr": 0.0001641931770763101, "epoch": 0.9140369967355821, "percentage": 30.47, "elapsed_time": "0:54:21", "remaining_time": "2:04:02"}
+{"current_steps": 850, "total_steps": 2757, "loss": 0.08000438213348389, "lr": 0.00016328210468285911, "epoch": 0.9249183895538629, "percentage": 30.83, "elapsed_time": "0:54:58", "remaining_time": "2:03:20"}
+{"current_steps": 860, "total_steps": 2757, "loss": 0.08225069642066955, "lr": 0.00016236218536143013, "epoch": 0.9357997823721437, "percentage": 31.19, "elapsed_time": "0:55:36", "remaining_time": "2:02:39"}
+{"current_steps": 870, "total_steps": 2757, "loss": 0.07968268394470215, "lr": 0.0001614335477180488, "epoch": 0.9466811751904244, "percentage": 31.56, "elapsed_time": "0:56:14", "remaining_time": "2:01:58"}
+{"current_steps": 880, "total_steps": 2757, "loss": 0.08057799339294433, "lr": 0.00016049632157757473, "epoch": 0.9575625680087051, "percentage": 31.92, "elapsed_time": "0:56:51", "remaining_time": "2:01:17"}
+{"current_steps": 890, "total_steps": 2757, "loss": 0.07912254929542542, "lr": 0.00015955063796555137, "epoch": 0.9684439608269858, "percentage": 32.28, "elapsed_time": "0:57:29", "remaining_time": "2:00:36"}
+{"current_steps": 900, "total_steps": 2757, "loss": 0.07660449743270874, "lr": 0.00015859662908988885, "epoch": 0.9793253536452666, "percentage": 32.64, "elapsed_time": "0:58:07", "remaining_time": "1:59:54"}
+{"current_steps": 910, "total_steps": 2757, "loss": 0.07685725092887878, "lr": 0.0001576344283223809, "epoch": 0.9902067464635473, "percentage": 33.01, "elapsed_time": "0:58:44", "remaining_time": "1:59:13"}
+{"current_steps": 920, "total_steps": 2757, "loss": 0.07933401465415954, "lr": 0.00015666417018005938, "epoch": 1.001088139281828, "percentage": 33.37, "elapsed_time": "0:59:22", "remaining_time": "1:58:33"}
+{"current_steps": 930, "total_steps": 2757, "loss": 0.07339736819267273, "lr": 0.00015568599030638854, "epoch": 1.0119695321001088, "percentage": 33.73, "elapsed_time": "0:59:59", "remaining_time": "1:57:51"}
+{"current_steps": 940, "total_steps": 2757, "loss": 0.07634537220001221, "lr": 0.00015470002545230193, "epoch": 1.0228509249183895, "percentage": 34.1, "elapsed_time": "1:00:39", "remaining_time": "1:57:14"}
+{"current_steps": 950, "total_steps": 2757, "loss": 0.08004393577575683, "lr": 0.0001537064134570844, "epoch": 1.0337323177366704, "percentage": 34.46, "elapsed_time": "1:01:17", "remaining_time": "1:56:34"}
+{"current_steps": 960, "total_steps": 2757, "loss": 0.07677503824234008, "lr": 0.0001527052932291019, "epoch": 1.0446137105549511, "percentage": 34.82, "elapsed_time": "1:01:54", "remaining_time": "1:55:52"}
+{"current_steps": 970, "total_steps": 2757, "loss": 0.07567052841186524, "lr": 0.00015169680472638218, "epoch": 1.0554951033732318, "percentage": 35.18, "elapsed_time": "1:02:31", "remaining_time": "1:55:11"}
+{"current_steps": 980, "total_steps": 2757, "loss": 0.0768008291721344, "lr": 0.00015068108893704797, "epoch": 1.0663764961915125, "percentage": 35.55, "elapsed_time": "1:03:08", "remaining_time": "1:54:30"}
+{"current_steps": 990, "total_steps": 2757, "loss": 0.07604062557220459, "lr": 0.0001496582878596071, "epoch": 1.0772578890097932, "percentage": 35.91, "elapsed_time": "1:03:46", "remaining_time": "1:53:49"}
+{"current_steps": 1000, "total_steps": 2757, "loss": 0.07631718516349792, "lr": 0.00014862854448310074, "epoch": 1.088139281828074, "percentage": 36.27, "elapsed_time": "1:04:23", "remaining_time": "1:53:08"}
+{"current_steps": 1000, "total_steps": 2757, "eval_loss": 0.08224387466907501, "epoch": 1.088139281828074, "percentage": 36.27, "elapsed_time": "1:04:29", "remaining_time": "1:53:18"}
+{"current_steps": 1010, "total_steps": 2757, "loss": 0.0752741813659668, "lr": 0.00014759200276711316, "epoch": 1.0990206746463547, "percentage": 36.63, "elapsed_time": "1:05:15", "remaining_time": "1:52:52"}
+{"current_steps": 1020, "total_steps": 2757, "loss": 0.07320383191108704, "lr": 0.00014654880762164624, "epoch": 1.1099020674646354, "percentage": 37.0, "elapsed_time": "1:05:52", "remaining_time": "1:52:10"}
+{"current_steps": 1030, "total_steps": 2757, "loss": 0.08079155683517455, "lr": 0.00014549910488686078, "epoch": 1.1207834602829163, "percentage": 37.36, "elapsed_time": "1:06:29", "remaining_time": "1:51:29"}
+{"current_steps": 1040, "total_steps": 2757, "loss": 0.07409957647323609, "lr": 0.0001444430413126877, "epoch": 1.131664853101197, "percentage": 37.72, "elapsed_time": "1:07:07", "remaining_time": "1:50:48"}
+{"current_steps": 1050, "total_steps": 2757, "loss": 0.07469584345817566, "lr": 0.0001433807645383123, "epoch": 1.1425462459194777, "percentage": 38.08, "elapsed_time": "1:07:44", "remaining_time": "1:50:07"}
+{"current_steps": 1060, "total_steps": 2757, "loss": 0.07519133687019348, "lr": 0.00014231242307153417, "epoch": 1.1534276387377584, "percentage": 38.45, "elapsed_time": "1:08:21", "remaining_time": "1:49:26"}
+{"current_steps": 1070, "total_steps": 2757, "loss": 0.07806066274642945, "lr": 0.00014123816626800535, "epoch": 1.1643090315560392, "percentage": 38.81, "elapsed_time": "1:08:59", "remaining_time": "1:48:46"}
+{"current_steps": 1080, "total_steps": 2757, "loss": 0.07402833104133606, "lr": 0.00014015814431035046, "epoch": 1.1751904243743199, "percentage": 39.17, "elapsed_time": "1:09:39", "remaining_time": "1:48:09"}
+{"current_steps": 1090, "total_steps": 2757, "loss": 0.07758893370628357, "lr": 0.00013907250818717082, "epoch": 1.1860718171926006, "percentage": 39.54, "elapsed_time": "1:10:16", "remaining_time": "1:47:29"}
+{"current_steps": 1100, "total_steps": 2757, "loss": 0.0765085220336914, "lr": 0.00013798140967193607, "epoch": 1.1969532100108813, "percentage": 39.9, "elapsed_time": "1:10:55", "remaining_time": "1:46:49"}
+{"current_steps": 1110, "total_steps": 2757, "loss": 0.07446168661117554, "lr": 0.0001368850013017659, "epoch": 1.2078346028291622, "percentage": 40.26, "elapsed_time": "1:11:32", "remaining_time": "1:46:09"}
+{"current_steps": 1120, "total_steps": 2757, "loss": 0.07697376012802123, "lr": 0.0001357834363561054, "epoch": 1.218715995647443, "percentage": 40.62, "elapsed_time": "1:12:10", "remaining_time": "1:45:29"}
+{"current_steps": 1130, "total_steps": 2757, "loss": 0.07688432335853576, "lr": 0.00013467686883529616, "epoch": 1.2295973884657236, "percentage": 40.99, "elapsed_time": "1:12:48", "remaining_time": "1:44:49"}
+{"current_steps": 1140, "total_steps": 2757, "loss": 0.07023751735687256, "lr": 0.0001335654534390468, "epoch": 1.2404787812840044, "percentage": 41.35, "elapsed_time": "1:13:25", "remaining_time": "1:44:09"}
+{"current_steps": 1150, "total_steps": 2757, "loss": 0.0761339783668518, "lr": 0.0001324493455448059, "epoch": 1.251360174102285, "percentage": 41.71, "elapsed_time": "1:14:03", "remaining_time": "1:43:29"}
+{"current_steps": 1160, "total_steps": 2757, "loss": 0.07061100006103516, "lr": 0.0001313287011860399, "epoch": 1.2622415669205658, "percentage": 42.07, "elapsed_time": "1:14:41", "remaining_time": "1:42:49"}
+{"current_steps": 1170, "total_steps": 2757, "loss": 0.07600026726722717, "lr": 0.00013020367703041934, "epoch": 1.2731229597388465, "percentage": 42.44, "elapsed_time": "1:15:18", "remaining_time": "1:42:09"}
+{"current_steps": 1180, "total_steps": 2757, "loss": 0.07545133829116821, "lr": 0.0001290744303579167, "epoch": 1.2840043525571274, "percentage": 42.8, "elapsed_time": "1:15:56", "remaining_time": "1:41:29"}
+{"current_steps": 1190, "total_steps": 2757, "loss": 0.07300611138343811, "lr": 0.00012794111903881818, "epoch": 1.2948857453754081, "percentage": 43.16, "elapsed_time": "1:16:33", "remaining_time": "1:40:49"}
+{"current_steps": 1200, "total_steps": 2757, "loss": 0.07233045101165772, "lr": 0.00012680390151165337, "epoch": 1.3057671381936888, "percentage": 43.53, "elapsed_time": "1:17:11", "remaining_time": "1:40:09"}
+{"current_steps": 1200, "total_steps": 2757, "eval_loss": 0.08009672164916992, "epoch": 1.3057671381936888, "percentage": 43.53, "elapsed_time": "1:17:17", "remaining_time": "1:40:16"}
+{"current_steps": 1210, "total_steps": 2757, "loss": 0.0722274899482727, "lr": 0.0001256629367610454, "epoch": 1.3166485310119695, "percentage": 43.89, "elapsed_time": "1:18:09", "remaining_time": "1:39:55"}
+{"current_steps": 1220, "total_steps": 2757, "loss": 0.07423741221427918, "lr": 0.0001245183842954844, "epoch": 1.3275299238302503, "percentage": 44.25, "elapsed_time": "1:18:47", "remaining_time": "1:39:15"}
+{"current_steps": 1230, "total_steps": 2757, "loss": 0.07771967649459839, "lr": 0.00012337040412502823, "epoch": 1.338411316648531, "percentage": 44.61, "elapsed_time": "1:19:25", "remaining_time": "1:38:35"}
+{"current_steps": 1240, "total_steps": 2757, "loss": 0.07685770392417908, "lr": 0.00012221915673893267, "epoch": 1.3492927094668117, "percentage": 44.98, "elapsed_time": "1:20:02", "remaining_time": "1:37:55"}
+{"current_steps": 1250, "total_steps": 2757, "loss": 0.0744680941104889, "lr": 0.00012106480308321498, "epoch": 1.3601741022850926, "percentage": 45.34, "elapsed_time": "1:20:40", "remaining_time": "1:37:15"}
+{"current_steps": 1260, "total_steps": 2757, "loss": 0.07823204398155212, "lr": 0.00011990750453815324, "epoch": 1.371055495103373, "percentage": 45.7, "elapsed_time": "1:21:19", "remaining_time": "1:36:37"}
+{"current_steps": 1270, "total_steps": 2757, "loss": 0.06924179792404175, "lr": 0.00011874742289572531, "epoch": 1.381936887921654, "percentage": 46.06, "elapsed_time": "1:21:57", "remaining_time": "1:35:58"}
+{"current_steps": 1280, "total_steps": 2757, "loss": 0.06726616621017456, "lr": 0.00011758472033698993, "epoch": 1.3928182807399347, "percentage": 46.43, "elapsed_time": "1:22:35", "remaining_time": "1:35:18"}
+{"current_steps": 1290, "total_steps": 2757, "loss": 0.07383161187171935, "lr": 0.00011641955940941374, "epoch": 1.4036996735582155, "percentage": 46.79, "elapsed_time": "1:23:13", "remaining_time": "1:34:38"}
+{"current_steps": 1300, "total_steps": 2757, "loss": 0.07341222763061524, "lr": 0.00011525210300414673, "epoch": 1.4145810663764962, "percentage": 47.15, "elapsed_time": "1:23:51", "remaining_time": "1:33:58"}
+{"current_steps": 1310, "total_steps": 2757, "loss": 0.07354844808578491, "lr": 0.00011408251433325006, "epoch": 1.4254624591947769, "percentage": 47.52, "elapsed_time": "1:24:28", "remaining_time": "1:33:19"}
+{"current_steps": 1320, "total_steps": 2757, "loss": 0.07191305160522461, "lr": 0.00011291095690687847, "epoch": 1.4363438520130578, "percentage": 47.88, "elapsed_time": "1:25:06", "remaining_time": "1:32:39"}
+{"current_steps": 1330, "total_steps": 2757, "loss": 0.07049686908721924, "lr": 0.00011173759451042168, "epoch": 1.4472252448313383, "percentage": 48.24, "elapsed_time": "1:25:43", "remaining_time": "1:31:58"}
+{"current_steps": 1340, "total_steps": 2757, "loss": 0.07753136157989501, "lr": 0.0001105625911816067, "epoch": 1.4581066376496192, "percentage": 48.6, "elapsed_time": "1:26:20", "remaining_time": "1:31:18"}
+{"current_steps": 1350, "total_steps": 2757, "loss": 0.07463480234146118, "lr": 0.0001093861111875652, "epoch": 1.4689880304679, "percentage": 48.97, "elapsed_time": "1:26:58", "remaining_time": "1:30:38"}
+{"current_steps": 1360, "total_steps": 2757, "loss": 0.06845179796218873, "lr": 0.0001082083190018688, "epoch": 1.4798694232861807, "percentage": 49.33, "elapsed_time": "1:27:35", "remaining_time": "1:29:58"}
+{"current_steps": 1370, "total_steps": 2757, "loss": 0.07124170064926147, "lr": 0.00010702937928153533, "epoch": 1.4907508161044614, "percentage": 49.69, "elapsed_time": "1:28:12", "remaining_time": "1:29:18"}
+{"current_steps": 1380, "total_steps": 2757, "loss": 0.07157056331634522, "lr": 0.00010584945684400963, "epoch": 1.501632208922742, "percentage": 50.05, "elapsed_time": "1:28:50", "remaining_time": "1:28:38"}
+{"current_steps": 1390, "total_steps": 2757, "loss": 0.07220081686973571, "lr": 0.0001046687166441217, "epoch": 1.512513601741023, "percentage": 50.42, "elapsed_time": "1:29:28", "remaining_time": "1:28:00"}
+{"current_steps": 1400, "total_steps": 2757, "loss": 0.07243863940238952, "lr": 0.00010348732375102601, "epoch": 1.5233949945593035, "percentage": 50.78, "elapsed_time": "1:30:07", "remaining_time": "1:27:21"}
+{"current_steps": 1400, "total_steps": 2757, "eval_loss": 0.07733941823244095, "epoch": 1.5233949945593035, "percentage": 50.78, "elapsed_time": "1:30:13", "remaining_time": "1:27:26"}
+{"current_steps": 1410, "total_steps": 2757, "loss": 0.06972206830978393, "lr": 0.00010230544332512442, "epoch": 1.5342763873775844, "percentage": 51.14, "elapsed_time": "1:31:00", "remaining_time": "1:26:56"}
+{"current_steps": 1420, "total_steps": 2757, "loss": 0.07575106620788574, "lr": 0.00010112324059497649, "epoch": 1.545157780195865, "percentage": 51.51, "elapsed_time": "1:31:37", "remaining_time": "1:26:16"}
+{"current_steps": 1430, "total_steps": 2757, "loss": 0.07188562154769898, "lr": 9.994088083420044e-05, "epoch": 1.5560391730141458, "percentage": 51.87, "elapsed_time": "1:32:14", "remaining_time": "1:25:36"}
+{"current_steps": 1440, "total_steps": 2757, "loss": 0.0751798152923584, "lr": 9.875852933836765e-05, "epoch": 1.5669205658324266, "percentage": 52.23, "elapsed_time": "1:32:52", "remaining_time": "1:24:56"}
+{"current_steps": 1450, "total_steps": 2757, "loss": 0.06870698928833008, "lr": 9.757635140189394e-05, "epoch": 1.5778019586507073, "percentage": 52.59, "elapsed_time": "1:33:30", "remaining_time": "1:24:16"}
+{"current_steps": 1460, "total_steps": 2757, "loss": 0.07310816049575805, "lr": 9.639451229493133e-05, "epoch": 1.588683351468988, "percentage": 52.96, "elapsed_time": "1:34:07", "remaining_time": "1:23:37"}
+{"current_steps": 1470, "total_steps": 2757, "loss": 0.07275789976119995, "lr": 9.521317724026299e-05, "epoch": 1.5995647442872687, "percentage": 53.32, "elapsed_time": "1:34:45", "remaining_time": "1:22:57"}
+{"current_steps": 1480, "total_steps": 2757, "loss": 0.07513962984085083, "lr": 9.403251139020486e-05, "epoch": 1.6104461371055496, "percentage": 53.68, "elapsed_time": "1:35:22", "remaining_time": "1:22:17"}
+{"current_steps": 1490, "total_steps": 2757, "loss": 0.0734050989151001, "lr": 9.285267980351719e-05, "epoch": 1.6213275299238301, "percentage": 54.04, "elapsed_time": "1:36:00", "remaining_time": "1:21:38"}
+{"current_steps": 1500, "total_steps": 2757, "loss": 0.06607317924499512, "lr": 9.167384742232898e-05, "epoch": 1.632208922742111, "percentage": 54.41, "elapsed_time": "1:36:38", "remaining_time": "1:20:58"}
+{"current_steps": 1510, "total_steps": 2757, "loss": 0.07146948575973511, "lr": 9.049617904907877e-05, "epoch": 1.6430903155603918, "percentage": 54.77, "elapsed_time": "1:37:15", "remaining_time": "1:20:19"}
+{"current_steps": 1520, "total_steps": 2757, "loss": 0.07334067225456238, "lr": 8.931983932347533e-05, "epoch": 1.6539717083786725, "percentage": 55.13, "elapsed_time": "1:37:52", "remaining_time": "1:19:39"}
+{"current_steps": 1530, "total_steps": 2757, "loss": 0.0697287678718567, "lr": 8.814499269948053e-05, "epoch": 1.6648531011969532, "percentage": 55.5, "elapsed_time": "1:38:30", "remaining_time": "1:18:59"}
+{"current_steps": 1540, "total_steps": 2757, "loss": 0.07186954021453858, "lr": 8.697180342231849e-05, "epoch": 1.6757344940152339, "percentage": 55.86, "elapsed_time": "1:39:07", "remaining_time": "1:18:19"}
+{"current_steps": 1550, "total_steps": 2757, "loss": 0.06877661943435669, "lr": 8.580043550551393e-05, "epoch": 1.6866158868335148, "percentage": 56.22, "elapsed_time": "1:39:45", "remaining_time": "1:17:41"}
+{"current_steps": 1560, "total_steps": 2757, "loss": 0.07183843851089478, "lr": 8.463105270796276e-05, "epoch": 1.6974972796517953, "percentage": 56.58, "elapsed_time": "1:40:23", "remaining_time": "1:17:02"}
+{"current_steps": 1570, "total_steps": 2757, "loss": 0.07082110047340393, "lr": 8.346381851103842e-05, "epoch": 1.7083786724700762, "percentage": 56.95, "elapsed_time": "1:41:01", "remaining_time": "1:16:22"}
+{"current_steps": 1580, "total_steps": 2757, "loss": 0.06895624399185181, "lr": 8.229889609573699e-05, "epoch": 1.719260065288357, "percentage": 57.31, "elapsed_time": "1:41:38", "remaining_time": "1:15:43"}
+{"current_steps": 1590, "total_steps": 2757, "loss": 0.0707672655582428, "lr": 8.113644831986416e-05, "epoch": 1.7301414581066377, "percentage": 57.67, "elapsed_time": "1:42:16", "remaining_time": "1:15:03"}
+{"current_steps": 1600, "total_steps": 2757, "loss": 0.07331972122192383, "lr": 7.997663769526753e-05, "epoch": 1.7410228509249184, "percentage": 58.03, "elapsed_time": "1:42:54", "remaining_time": "1:14:24"}
+{"current_steps": 1600, "total_steps": 2757, "eval_loss": 0.07515527307987213, "epoch": 1.7410228509249184, "percentage": 58.03, "elapsed_time": "1:42:59", "remaining_time": "1:14:28"}
+{"current_steps": 1610, "total_steps": 2757, "loss": 0.07261155843734741, "lr": 7.881962636511715e-05, "epoch": 1.751904243743199, "percentage": 58.4, "elapsed_time": "1:43:50", "remaining_time": "1:13:58"}
+{"current_steps": 1620, "total_steps": 2757, "loss": 0.07076815962791443, "lr": 7.766557608123783e-05, "epoch": 1.76278563656148, "percentage": 58.76, "elapsed_time": "1:44:28", "remaining_time": "1:13:19"}
+{"current_steps": 1630, "total_steps": 2757, "loss": 0.0643875777721405, "lr": 7.651464818149587e-05, "epoch": 1.7736670293797605, "percentage": 59.12, "elapsed_time": "1:45:05", "remaining_time": "1:12:39"}
+{"current_steps": 1640, "total_steps": 2757, "loss": 0.07072726488113404, "lr": 7.536700356724379e-05, "epoch": 1.7845484221980414, "percentage": 59.48, "elapsed_time": "1:45:44", "remaining_time": "1:12:00"}
+{"current_steps": 1650, "total_steps": 2757, "loss": 0.07485873699188232, "lr": 7.422280268082629e-05, "epoch": 1.795429815016322, "percentage": 59.85, "elapsed_time": "1:46:21", "remaining_time": "1:11:21"}
+{"current_steps": 1660, "total_steps": 2757, "loss": 0.06636062860488892, "lr": 7.308220548314981e-05, "epoch": 1.8063112078346029, "percentage": 60.21, "elapsed_time": "1:47:00", "remaining_time": "1:10:43"}
+{"current_steps": 1670, "total_steps": 2757, "loss": 0.06702766418457032, "lr": 7.194537143132023e-05, "epoch": 1.8171926006528836, "percentage": 60.57, "elapsed_time": "1:47:38", "remaining_time": "1:10:03"}
+{"current_steps": 1680, "total_steps": 2757, "loss": 0.07087734937667847, "lr": 7.081245945635013e-05, "epoch": 1.8280739934711643, "percentage": 60.94, "elapsed_time": "1:48:16", "remaining_time": "1:09:24"}
+{"current_steps": 1690, "total_steps": 2757, "loss": 0.06891222596168518, "lr": 6.968362794094027e-05, "epoch": 1.8389553862894452, "percentage": 61.3, "elapsed_time": "1:48:54", "remaining_time": "1:08:45"}
+{"current_steps": 1700, "total_steps": 2757, "loss": 0.06720860004425049, "lr": 6.855903469733738e-05, "epoch": 1.8498367791077257, "percentage": 61.66, "elapsed_time": "1:49:31", "remaining_time": "1:08:06"}
+{"current_steps": 1710, "total_steps": 2757, "loss": 0.06776891946792603, "lr": 6.743883694527169e-05, "epoch": 1.8607181719260066, "percentage": 62.02, "elapsed_time": "1:50:10", "remaining_time": "1:07:27"}
+{"current_steps": 1720, "total_steps": 2757, "loss": 0.07007836103439331, "lr": 6.632319128997759e-05, "epoch": 1.8715995647442871, "percentage": 62.39, "elapsed_time": "1:50:47", "remaining_time": "1:06:48"}
+{"current_steps": 1730, "total_steps": 2757, "loss": 0.07095590829849244, "lr": 6.52122537002998e-05, "epoch": 1.882480957562568, "percentage": 62.75, "elapsed_time": "1:51:25", "remaining_time": "1:06:08"}
+{"current_steps": 1740, "total_steps": 2757, "loss": 0.06855381727218628, "lr": 6.410617948688892e-05, "epoch": 1.8933623503808488, "percentage": 63.11, "elapsed_time": "1:52:03", "remaining_time": "1:05:29"}
+{"current_steps": 1750, "total_steps": 2757, "loss": 0.07089903354644775, "lr": 6.300512328048845e-05, "epoch": 1.9042437431991295, "percentage": 63.47, "elapsed_time": "1:52:40", "remaining_time": "1:04:50"}
+{"current_steps": 1760, "total_steps": 2757, "loss": 0.0708732008934021, "lr": 6.190923901031757e-05, "epoch": 1.9151251360174102, "percentage": 63.84, "elapsed_time": "1:53:17", "remaining_time": "1:04:10"}
+{"current_steps": 1770, "total_steps": 2757, "loss": 0.07153573036193847, "lr": 6.081867988255135e-05, "epoch": 1.926006528835691, "percentage": 64.2, "elapsed_time": "1:53:55", "remaining_time": "1:03:31"}
+{"current_steps": 1780, "total_steps": 2757, "loss": 0.0697596788406372, "lr": 5.973359835890246e-05, "epoch": 1.9368879216539718, "percentage": 64.56, "elapsed_time": "1:54:32", "remaining_time": "1:02:52"}
+{"current_steps": 1790, "total_steps": 2757, "loss": 0.06653320789337158, "lr": 5.865414613530669e-05, "epoch": 1.9477693144722523, "percentage": 64.93, "elapsed_time": "1:55:09", "remaining_time": "1:02:12"}
+{"current_steps": 1800, "total_steps": 2757, "loss": 0.07076951861381531, "lr": 5.758047412071568e-05, "epoch": 1.9586507072905333, "percentage": 65.29, "elapsed_time": "1:55:46", "remaining_time": "1:01:33"}
+{"current_steps": 1800, "total_steps": 2757, "eval_loss": 0.07202806323766708, "epoch": 1.9586507072905333, "percentage": 65.29, "elapsed_time": "1:55:52", "remaining_time": "1:01:36"}
+{"current_steps": 1810, "total_steps": 2757, "loss": 0.06910277605056762, "lr": 5.651273241599976e-05, "epoch": 1.969532100108814, "percentage": 65.65, "elapsed_time": "1:56:38", "remaining_time": "1:01:01"}
+{"current_steps": 1820, "total_steps": 2757, "loss": 0.06721980571746826, "lr": 5.5451070292963434e-05, "epoch": 1.9804134929270947, "percentage": 66.01, "elapsed_time": "1:57:15", "remaining_time": "1:00:22"}
+{"current_steps": 1830, "total_steps": 2757, "loss": 0.06764072179794312, "lr": 5.439563617347711e-05, "epoch": 1.9912948857453754, "percentage": 66.38, "elapsed_time": "1:57:52", "remaining_time": "0:59:42"}
+{"current_steps": 1840, "total_steps": 2757, "loss": 0.07124176025390624, "lr": 5.3346577608727463e-05, "epoch": 2.002176278563656, "percentage": 66.74, "elapsed_time": "1:58:30", "remaining_time": "0:59:03"}
+{"current_steps": 1850, "total_steps": 2757, "loss": 0.05957789421081543, "lr": 5.230404125858952e-05, "epoch": 2.013057671381937, "percentage": 67.1, "elapsed_time": "1:59:07", "remaining_time": "0:58:24"}
+{"current_steps": 1860, "total_steps": 2757, "loss": 0.055424296855926515, "lr": 5.1268172871123535e-05, "epoch": 2.0239390642002175, "percentage": 67.46, "elapsed_time": "1:59:44", "remaining_time": "0:57:44"}
+{"current_steps": 1870, "total_steps": 2757, "loss": 0.06114422678947449, "lr": 5.023911726219902e-05, "epoch": 2.0348204570184985, "percentage": 67.83, "elapsed_time": "2:00:21", "remaining_time": "0:57:05"}
+{"current_steps": 1880, "total_steps": 2757, "loss": 0.058832204341888426, "lr": 4.921701829524956e-05, "epoch": 2.045701849836779, "percentage": 68.19, "elapsed_time": "2:00:59", "remaining_time": "0:56:26"}
+{"current_steps": 1890, "total_steps": 2757, "loss": 0.057713437080383304, "lr": 4.820201886116022e-05, "epoch": 2.05658324265506, "percentage": 68.55, "elapsed_time": "2:01:36", "remaining_time": "0:55:46"}
+{"current_steps": 1900, "total_steps": 2757, "loss": 0.05968807935714722, "lr": 4.719426085829123e-05, "epoch": 2.067464635473341, "percentage": 68.92, "elapsed_time": "2:02:13", "remaining_time": "0:55:07"}
+{"current_steps": 1910, "total_steps": 2757, "loss": 0.060094785690307614, "lr": 4.619388517264056e-05, "epoch": 2.0783460282916213, "percentage": 69.28, "elapsed_time": "2:02:50", "remaining_time": "0:54:28"}
+{"current_steps": 1920, "total_steps": 2757, "loss": 0.05601954460144043, "lr": 4.5201031658147766e-05, "epoch": 2.0892274211099022, "percentage": 69.64, "elapsed_time": "2:03:27", "remaining_time": "0:53:49"}
+{"current_steps": 1930, "total_steps": 2757, "loss": 0.06034954190254212, "lr": 4.4215839117142134e-05, "epoch": 2.1001088139281827, "percentage": 70.0, "elapsed_time": "2:04:05", "remaining_time": "0:53:10"}
+{"current_steps": 1940, "total_steps": 2757, "loss": 0.05510505437850952, "lr": 4.3238445280938154e-05, "epoch": 2.1109902067464636, "percentage": 70.37, "elapsed_time": "2:04:43", "remaining_time": "0:52:31"}
+{"current_steps": 1950, "total_steps": 2757, "loss": 0.060755813121795656, "lr": 4.2268986790580125e-05, "epoch": 2.121871599564744, "percentage": 70.73, "elapsed_time": "2:05:21", "remaining_time": "0:51:52"}
+{"current_steps": 1960, "total_steps": 2757, "loss": 0.06080647706985474, "lr": 4.130759917773985e-05, "epoch": 2.132752992383025, "percentage": 71.09, "elapsed_time": "2:05:58", "remaining_time": "0:51:13"}
+{"current_steps": 1970, "total_steps": 2757, "loss": 0.06485359072685241, "lr": 4.0354416845769e-05, "epoch": 2.143634385201306, "percentage": 71.45, "elapsed_time": "2:06:36", "remaining_time": "0:50:34"}
+{"current_steps": 1980, "total_steps": 2757, "loss": 0.059920203685760495, "lr": 3.940957305090935e-05, "epoch": 2.1545157780195865, "percentage": 71.82, "elapsed_time": "2:07:13", "remaining_time": "0:49:55"}
+{"current_steps": 1990, "total_steps": 2757, "loss": 0.0490723729133606, "lr": 3.8473199883663334e-05, "epoch": 2.1653971708378674, "percentage": 72.18, "elapsed_time": "2:07:51", "remaining_time": "0:49:16"}
+{"current_steps": 2000, "total_steps": 2757, "loss": 0.053455913066864015, "lr": 3.754542825032766e-05, "epoch": 2.176278563656148, "percentage": 72.54, "elapsed_time": "2:08:28", "remaining_time": "0:48:37"}
+{"current_steps": 2000, "total_steps": 2757, "eval_loss": 0.07033731788396835, "epoch": 2.176278563656148, "percentage": 72.54, "elapsed_time": "2:08:33", "remaining_time": "0:48:39"}
+{"current_steps": 2010, "total_steps": 2757, "loss": 0.061966001987457275, "lr": 3.6626387854692365e-05, "epoch": 2.187159956474429, "percentage": 72.91, "elapsed_time": "2:09:20", "remaining_time": "0:48:04"}
+{"current_steps": 2020, "total_steps": 2757, "loss": 0.05781112909317017, "lr": 3.571620717990811e-05, "epoch": 2.1980413492927093, "percentage": 73.27, "elapsed_time": "2:09:57", "remaining_time": "0:47:25"}
+{"current_steps": 2030, "total_steps": 2757, "loss": 0.060006117820739745, "lr": 3.4815013470524004e-05, "epoch": 2.2089227421109903, "percentage": 73.63, "elapsed_time": "2:10:35", "remaining_time": "0:46:45"}
+{"current_steps": 2040, "total_steps": 2757, "loss": 0.05959203839302063, "lr": 3.392293271469866e-05, "epoch": 2.2198041349292708, "percentage": 73.99, "elapsed_time": "2:11:12", "remaining_time": "0:46:06"}
+{"current_steps": 2050, "total_steps": 2757, "loss": 0.057662928104400636, "lr": 3.304008962658677e-05, "epoch": 2.2306855277475517, "percentage": 74.36, "elapsed_time": "2:11:49", "remaining_time": "0:45:27"}
+{"current_steps": 2060, "total_steps": 2757, "loss": 0.05794321894645691, "lr": 3.216660762890413e-05, "epoch": 2.2415669205658326, "percentage": 74.72, "elapsed_time": "2:12:27", "remaining_time": "0:44:49"}
+{"current_steps": 2070, "total_steps": 2757, "loss": 0.05402252674102783, "lr": 3.130260883567272e-05, "epoch": 2.252448313384113, "percentage": 75.08, "elapsed_time": "2:13:05", "remaining_time": "0:44:10"}
+{"current_steps": 2080, "total_steps": 2757, "loss": 0.06356992721557617, "lr": 3.0448214035149102e-05, "epoch": 2.263329706202394, "percentage": 75.44, "elapsed_time": "2:13:42", "remaining_time": "0:43:31"}
+{"current_steps": 2090, "total_steps": 2757, "loss": 0.057219624519348145, "lr": 2.960354267293809e-05, "epoch": 2.2742110990206745, "percentage": 75.81, "elapsed_time": "2:14:19", "remaining_time": "0:42:52"}
+{"current_steps": 2100, "total_steps": 2757, "loss": 0.0654039740562439, "lr": 2.8768712835293922e-05, "epoch": 2.2850924918389555, "percentage": 76.17, "elapsed_time": "2:14:57", "remaining_time": "0:42:13"}
+{"current_steps": 2110, "total_steps": 2757, "loss": 0.05935961604118347, "lr": 2.7943841232611812e-05, "epoch": 2.295973884657236, "percentage": 76.53, "elapsed_time": "2:15:35", "remaining_time": "0:41:34"}
+{"current_steps": 2120, "total_steps": 2757, "loss": 0.059474557638168335, "lr": 2.712904318311167e-05, "epoch": 2.306855277475517, "percentage": 76.9, "elapsed_time": "2:16:12", "remaining_time": "0:40:55"}
+{"current_steps": 2130, "total_steps": 2757, "loss": 0.06448624134063721, "lr": 2.6324432596716276e-05, "epoch": 2.3177366702937974, "percentage": 77.26, "elapsed_time": "2:16:51", "remaining_time": "0:40:17"}
+{"current_steps": 2140, "total_steps": 2757, "loss": 0.055650746822357176, "lr": 2.5530121959126718e-05, "epoch": 2.3286180631120783, "percentage": 77.62, "elapsed_time": "2:17:28", "remaining_time": "0:39:38"}
+{"current_steps": 2150, "total_steps": 2757, "loss": 0.05966170430183411, "lr": 2.47462223160965e-05, "epoch": 2.3394994559303592, "percentage": 77.98, "elapsed_time": "2:18:06", "remaining_time": "0:38:59"}
+{"current_steps": 2160, "total_steps": 2757, "loss": 0.06053815484046936, "lr": 2.397284325790744e-05, "epoch": 2.3503808487486397, "percentage": 78.35, "elapsed_time": "2:18:43", "remaining_time": "0:38:20"}
+{"current_steps": 2170, "total_steps": 2757, "loss": 0.0560091495513916, "lr": 2.321009290404864e-05, "epoch": 2.3612622415669207, "percentage": 78.71, "elapsed_time": "2:19:21", "remaining_time": "0:37:41"}
+{"current_steps": 2180, "total_steps": 2757, "loss": 0.058753132820129395, "lr": 2.2458077888101226e-05, "epoch": 2.372143634385201, "percentage": 79.07, "elapsed_time": "2:19:59", "remaining_time": "0:37:03"}
+{"current_steps": 2190, "total_steps": 2757, "loss": 0.05801019072532654, "lr": 2.1716903342830995e-05, "epoch": 2.383025027203482, "percentage": 79.43, "elapsed_time": "2:20:37", "remaining_time": "0:36:24"}
+{"current_steps": 2200, "total_steps": 2757, "loss": 0.06117302179336548, "lr": 2.0986672885490378e-05, "epoch": 2.3939064200217626, "percentage": 79.8, "elapsed_time": "2:21:16", "remaining_time": "0:35:45"}
+{"current_steps": 2200, "total_steps": 2757, "eval_loss": 0.0698942095041275, "epoch": 2.3939064200217626, "percentage": 79.8, "elapsed_time": "2:21:21", "remaining_time": "0:35:47"}
+{"current_steps": 2210, "total_steps": 2757, "loss": 0.05466359853744507, "lr": 2.0267488603332907e-05, "epoch": 2.4047878128400435, "percentage": 80.16, "elapsed_time": "2:22:09", "remaining_time": "0:35:11"}
+{"current_steps": 2220, "total_steps": 2757, "loss": 0.05731322765350342, "lr": 1.9559451039341147e-05, "epoch": 2.4156692056583244, "percentage": 80.52, "elapsed_time": "2:22:46", "remaining_time": "0:34:32"}
+{"current_steps": 2230, "total_steps": 2757, "loss": 0.06014159917831421, "lr": 1.886265917817064e-05, "epoch": 2.426550598476605, "percentage": 80.89, "elapsed_time": "2:23:24", "remaining_time": "0:33:53"}
+{"current_steps": 2240, "total_steps": 2757, "loss": 0.05688718557357788, "lr": 1.8177210432311732e-05, "epoch": 2.437431991294886, "percentage": 81.25, "elapsed_time": "2:24:01", "remaining_time": "0:33:14"}
+{"current_steps": 2250, "total_steps": 2757, "loss": 0.058925163745880124, "lr": 1.750320062847116e-05, "epoch": 2.4483133841131663, "percentage": 81.61, "elapsed_time": "2:24:38", "remaining_time": "0:32:35"}
+{"current_steps": 2260, "total_steps": 2757, "loss": 0.05712746381759644, "lr": 1.6840723994175356e-05, "epoch": 2.4591947769314473, "percentage": 81.97, "elapsed_time": "2:25:16", "remaining_time": "0:31:56"}
+{"current_steps": 2270, "total_steps": 2757, "loss": 0.056266939640045165, "lr": 1.6189873144597267e-05, "epoch": 2.4700761697497278, "percentage": 82.34, "elapsed_time": "2:25:54", "remaining_time": "0:31:18"}
+{"current_steps": 2280, "total_steps": 2757, "loss": 0.05650065541267395, "lr": 1.5550739069608676e-05, "epoch": 2.4809575625680087, "percentage": 82.7, "elapsed_time": "2:26:31", "remaining_time": "0:30:39"}
+{"current_steps": 2290, "total_steps": 2757, "loss": 0.05500165820121765, "lr": 1.492341112105966e-05, "epoch": 2.4918389553862896, "percentage": 83.06, "elapsed_time": "2:27:09", "remaining_time": "0:30:00"}
+{"current_steps": 2300, "total_steps": 2757, "loss": 0.05969707369804382, "lr": 1.4307977000287064e-05, "epoch": 2.50272034820457, "percentage": 83.42, "elapsed_time": "2:27:46", "remaining_time": "0:29:21"}
+{"current_steps": 2310, "total_steps": 2757, "loss": 0.05166618824005127, "lr": 1.3704522745853743e-05, "epoch": 2.513601741022851, "percentage": 83.79, "elapsed_time": "2:28:23", "remaining_time": "0:28:42"}
+{"current_steps": 2320, "total_steps": 2757, "loss": 0.0567930817604065, "lr": 1.311313272152025e-05, "epoch": 2.5244831338411315, "percentage": 84.15, "elapsed_time": "2:29:01", "remaining_time": "0:28:04"}
+{"current_steps": 2330, "total_steps": 2757, "loss": 0.05777972936630249, "lr": 1.2533889604450677e-05, "epoch": 2.5353645266594125, "percentage": 84.51, "elapsed_time": "2:29:39", "remaining_time": "0:27:25"}
+{"current_steps": 2340, "total_steps": 2757, "loss": 0.054594647884368894, "lr": 1.196687437365427e-05, "epoch": 2.546245919477693, "percentage": 84.87, "elapsed_time": "2:30:17", "remaining_time": "0:26:46"}
+{"current_steps": 2350, "total_steps": 2757, "loss": 0.05491857528686524, "lr": 1.141216629866434e-05, "epoch": 2.557127312295974, "percentage": 85.24, "elapsed_time": "2:30:55", "remaining_time": "0:26:08"}
+{"current_steps": 2360, "total_steps": 2757, "loss": 0.05289023518562317, "lr": 1.0869842928456441e-05, "epoch": 2.568008705114255, "percentage": 85.6, "elapsed_time": "2:31:32", "remaining_time": "0:25:29"}
+{"current_steps": 2370, "total_steps": 2757, "loss": 0.05806971788406372, "lr": 1.033998008060687e-05, "epoch": 2.5788900979325353, "percentage": 85.96, "elapsed_time": "2:32:09", "remaining_time": "0:24:50"}
+{"current_steps": 2380, "total_steps": 2757, "loss": 0.05867668390274048, "lr": 9.822651830693152e-06, "epoch": 2.5897714907508163, "percentage": 86.33, "elapsed_time": "2:32:47", "remaining_time": "0:24:12"}
+{"current_steps": 2390, "total_steps": 2757, "loss": 0.05183411836624145, "lr": 9.317930501938289e-06, "epoch": 2.6006528835690967, "percentage": 86.69, "elapsed_time": "2:33:24", "remaining_time": "0:23:33"}
+{"current_steps": 2400, "total_steps": 2757, "loss": 0.06119837760925293, "lr": 8.82588665509978e-06, "epoch": 2.6115342763873777, "percentage": 87.05, "elapsed_time": "2:34:01", "remaining_time": "0:22:54"}
+{"current_steps": 2400, "total_steps": 2757, "eval_loss": 0.06727322936058044, "epoch": 2.6115342763873777, "percentage": 87.05, "elapsed_time": "2:34:07", "remaining_time": "0:22:55"}
+{"current_steps": 2410, "total_steps": 2757, "loss": 0.05637891292572021, "lr": 8.346589078605215e-06, "epoch": 2.622415669205658, "percentage": 87.41, "elapsed_time": "2:34:56", "remaining_time": "0:22:18"}
+{"current_steps": 2420, "total_steps": 2757, "loss": 0.05910442471504211, "lr": 7.880104778935505e-06, "epoch": 2.633297062023939, "percentage": 87.78, "elapsed_time": "2:35:34", "remaining_time": "0:21:39"}
+{"current_steps": 2430, "total_steps": 2757, "loss": 0.054590433835983276, "lr": 7.426498971257278e-06, "epoch": 2.64417845484222, "percentage": 88.14, "elapsed_time": "2:36:11", "remaining_time": "0:21:01"}
+{"current_steps": 2440, "total_steps": 2757, "loss": 0.056293058395385745, "lr": 6.985835070305768e-06, "epoch": 2.6550598476605005, "percentage": 88.5, "elapsed_time": "2:36:49", "remaining_time": "0:20:22"}
+{"current_steps": 2450, "total_steps": 2757, "loss": 0.058118653297424314, "lr": 6.5581746815192135e-06, "epoch": 2.6659412404787814, "percentage": 88.86, "elapsed_time": "2:37:26", "remaining_time": "0:19:43"}
+{"current_steps": 2460, "total_steps": 2757, "loss": 0.05446176528930664, "lr": 6.143577592426453e-06, "epoch": 2.676822633297062, "percentage": 89.23, "elapsed_time": "2:38:04", "remaining_time": "0:19:05"}
+{"current_steps": 2470, "total_steps": 2757, "loss": 0.05432036519050598, "lr": 5.742101764288521e-06, "epoch": 2.687704026115343, "percentage": 89.59, "elapsed_time": "2:38:42", "remaining_time": "0:18:26"}
+{"current_steps": 2480, "total_steps": 2757, "loss": 0.05071129202842713, "lr": 5.353803323995521e-06, "epoch": 2.6985854189336234, "percentage": 89.95, "elapsed_time": "2:39:19", "remaining_time": "0:17:47"}
+{"current_steps": 2490, "total_steps": 2757, "loss": 0.04942301511764526, "lr": 4.9787365562200985e-06, "epoch": 2.7094668117519043, "percentage": 90.32, "elapsed_time": "2:39:56", "remaining_time": "0:17:09"}
+{"current_steps": 2500, "total_steps": 2757, "loss": 0.05862914323806763, "lr": 4.616953895828269e-06, "epoch": 2.7203482045701852, "percentage": 90.68, "elapsed_time": "2:40:34", "remaining_time": "0:16:30"}
+{"current_steps": 2510, "total_steps": 2757, "loss": 0.054695284366607665, "lr": 4.268505920549016e-06, "epoch": 2.7312295973884657, "percentage": 91.04, "elapsed_time": "2:41:11", "remaining_time": "0:15:51"}
+{"current_steps": 2520, "total_steps": 2757, "loss": 0.04879617691040039, "lr": 3.933441343903444e-06, "epoch": 2.742110990206746, "percentage": 91.4, "elapsed_time": "2:41:49", "remaining_time": "0:15:13"}
+{"current_steps": 2530, "total_steps": 2757, "loss": 0.05482098460197449, "lr": 3.61180700839453e-06, "epoch": 2.752992383025027, "percentage": 91.77, "elapsed_time": "2:42:26", "remaining_time": "0:14:34"}
+{"current_steps": 2540, "total_steps": 2757, "loss": 0.05537814497947693, "lr": 3.3036478789585044e-06, "epoch": 2.763873775843308, "percentage": 92.13, "elapsed_time": "2:43:04", "remaining_time": "0:13:55"}
+{"current_steps": 2550, "total_steps": 2757, "loss": 0.05486249327659607, "lr": 3.0090070366786792e-06, "epoch": 2.7747551686615886, "percentage": 92.49, "elapsed_time": "2:43:41", "remaining_time": "0:13:17"}
+{"current_steps": 2560, "total_steps": 2757, "loss": 0.054778552055358885, "lr": 2.7279256727626546e-06, "epoch": 2.7856365614798695, "percentage": 92.85, "elapsed_time": "2:44:19", "remaining_time": "0:12:38"}
+{"current_steps": 2570, "total_steps": 2757, "loss": 0.05409228205680847, "lr": 2.4604430827837143e-06, "epoch": 2.7965179542981504, "percentage": 93.22, "elapsed_time": "2:44:56", "remaining_time": "0:12:00"}
+{"current_steps": 2580, "total_steps": 2757, "loss": 0.06026080846786499, "lr": 2.2065966611872657e-06, "epoch": 2.807399347116431, "percentage": 93.58, "elapsed_time": "2:45:33", "remaining_time": "0:11:21"}
+{"current_steps": 2590, "total_steps": 2757, "loss": 0.05407561063766479, "lr": 1.9664218960630533e-06, "epoch": 2.8182807399347114, "percentage": 93.94, "elapsed_time": "2:46:11", "remaining_time": "0:10:42"}
+{"current_steps": 2600, "total_steps": 2757, "loss": 0.0522362232208252, "lr": 1.7399523641838278e-06, "epoch": 2.8291621327529923, "percentage": 94.31, "elapsed_time": "2:46:48", "remaining_time": "0:10:04"}
+{"current_steps": 2600, "total_steps": 2757, "eval_loss": 0.06709529459476471, "epoch": 2.8291621327529923, "percentage": 94.31, "elapsed_time": "2:46:54", "remaining_time": "0:10:04"}
+{"current_steps": 2610, "total_steps": 2757, "loss": 0.050741416215896604, "lr": 1.5272197263113131e-06, "epoch": 2.8400435255712733, "percentage": 94.67, "elapsed_time": "2:47:42", "remaining_time": "0:09:26"}
+{"current_steps": 2620, "total_steps": 2757, "loss": 0.054420065879821775, "lr": 1.3282537227699787e-06, "epoch": 2.8509249183895538, "percentage": 95.03, "elapsed_time": "2:48:19", "remaining_time": "0:08:48"}
+{"current_steps": 2630, "total_steps": 2757, "loss": 0.05170666575431824, "lr": 1.143082169289289e-06, "epoch": 2.8618063112078347, "percentage": 95.39, "elapsed_time": "2:48:58", "remaining_time": "0:08:09"}
+{"current_steps": 2640, "total_steps": 2757, "loss": 0.05548094511032105, "lr": 9.717309531150132e-07, "epoch": 2.8726877040261156, "percentage": 95.76, "elapsed_time": "2:49:35", "remaining_time": "0:07:30"}
+{"current_steps": 2650, "total_steps": 2757, "loss": 0.05571337938308716, "lr": 8.142240293901649e-07, "epoch": 2.883569096844396, "percentage": 96.12, "elapsed_time": "2:50:13", "remaining_time": "0:06:52"}
+{"current_steps": 2660, "total_steps": 2757, "loss": 0.05904418230056763, "lr": 6.705834178060589e-07, "epoch": 2.8944504896626766, "percentage": 96.48, "elapsed_time": "2:50:50", "remaining_time": "0:06:13"}
+{"current_steps": 2670, "total_steps": 2757, "loss": 0.0517579436302185, "lr": 5.408291995238956e-07, "epoch": 2.9053318824809575, "percentage": 96.84, "elapsed_time": "2:51:28", "remaining_time": "0:05:35"}
+{"current_steps": 2680, "total_steps": 2757, "loss": 0.05420621633529663, "lr": 4.249795143673962e-07, "epoch": 2.9162132752992385, "percentage": 97.21, "elapsed_time": "2:52:06", "remaining_time": "0:04:56"}
+{"current_steps": 2690, "total_steps": 2757, "loss": 0.060290944576263425, "lr": 3.230505582868637e-07, "epoch": 2.927094668117519, "percentage": 97.57, "elapsed_time": "2:52:45", "remaining_time": "0:04:18"}
+{"current_steps": 2700, "total_steps": 2757, "loss": 0.05462511777877808, "lr": 2.3505658109487238e-07, "epoch": 2.9379760609358, "percentage": 97.93, "elapsed_time": "2:53:23", "remaining_time": "0:03:39"}
+{"current_steps": 2710, "total_steps": 2757, "loss": 0.06044437885284424, "lr": 1.6100988447423915e-07, "epoch": 2.9488574537540804, "percentage": 98.3, "elapsed_time": "2:54:04", "remaining_time": "0:03:01"}
+{"current_steps": 2720, "total_steps": 2757, "loss": 0.0504527747631073, "lr": 1.0092082025815464e-07, "epoch": 2.9597388465723613, "percentage": 98.66, "elapsed_time": "2:54:43", "remaining_time": "0:02:22"}
+{"current_steps": 2730, "total_steps": 2757, "loss": 0.053725385665893556, "lr": 5.479778898302979e-08, "epoch": 2.970620239390642, "percentage": 99.02, "elapsed_time": "2:55:21", "remaining_time": "0:01:44"}
+{"current_steps": 2740, "total_steps": 2757, "loss": 0.05662902593612671, "lr": 2.2647238714024188e-08, "epoch": 2.9815016322089227, "percentage": 99.38, "elapsed_time": "2:55:58", "remaining_time": "0:01:05"}
+{"current_steps": 2750, "total_steps": 2757, "loss": 0.0520307719707489, "lr": 4.473664143667122e-09, "epoch": 2.9923830250272037, "percentage": 99.75, "elapsed_time": "2:56:36", "remaining_time": "0:00:26"}
+{"current_steps": 2757, "total_steps": 2757, "epoch": 3.0, "percentage": 100.0, "elapsed_time": "2:57:11", "remaining_time": "0:00:00"}

qwen3_8b_sft_rtl/trainer_state.json

@@ -0,0 +1,2072 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 200,
+  "global_step": 2757,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.01088139281828074,
+      "grad_norm": 0.2948031425476074,
+      "learning_rate": 1.8e-05,
+      "loss": 0.741086769104004,
+      "step": 10
+    },
+    {
+      "epoch": 0.02176278563656148,
+      "grad_norm": 0.15493904054164886,
+      "learning_rate": 3.8e-05,
+      "loss": 0.3913477897644043,
+      "step": 20
+    },
+    {
+      "epoch": 0.03264417845484222,
+      "grad_norm": 0.136435404419899,
+      "learning_rate": 5.8e-05,
+      "loss": 0.22936077117919923,
+      "step": 30
+    },
+    {
+      "epoch": 0.04352557127312296,
+      "grad_norm": 0.09819520264863968,
+      "learning_rate": 7.800000000000001e-05,
+      "loss": 0.17587745189666748,
+      "step": 40
+    },
+    {
+      "epoch": 0.0544069640914037,
+      "grad_norm": 0.10766161978244781,
+      "learning_rate": 9.8e-05,
+      "loss": 0.14949668645858766,
+      "step": 50
+    },
+    {
+      "epoch": 0.06528835690968444,
+      "grad_norm": 0.08924636989831924,
+      "learning_rate": 0.000118,
+      "loss": 0.13353660106658935,
+      "step": 60
+    },
+    {
+      "epoch": 0.07616974972796518,
+      "grad_norm": 0.08617673069238663,
+      "learning_rate": 0.000138,
+      "loss": 0.120628023147583,
+      "step": 70
+    },
+    {
+      "epoch": 0.08705114254624592,
+      "grad_norm": 0.06150153651833534,
+      "learning_rate": 0.00015800000000000002,
+      "loss": 0.11396801471710205,
+      "step": 80
+    },
+    {
+      "epoch": 0.09793253536452666,
+      "grad_norm": 0.06216437742114067,
+      "learning_rate": 0.00017800000000000002,
+      "loss": 0.1074080228805542,
+      "step": 90
+    },
+    {
+      "epoch": 0.1088139281828074,
+      "grad_norm": 0.07203802466392517,
+      "learning_rate": 0.00019800000000000002,
+      "loss": 0.10922808647155761,
+      "step": 100
+    },
+    {
+      "epoch": 0.11969532100108814,
+      "grad_norm": 0.04227524250745773,
+      "learning_rate": 0.00019999433803003197,
+      "loss": 0.10778391361236572,
+      "step": 110
+    },
+    {
+      "epoch": 0.1305767138193689,
+      "grad_norm": 0.06469517946243286,
+      "learning_rate": 0.00019997476661131597,
+      "loss": 0.10001739263534545,
+      "step": 120
+    },
+    {
+      "epoch": 0.14145810663764963,
+      "grad_norm": 0.04663902893662453,
+      "learning_rate": 0.0001999412185784487,
+      "loss": 0.10446863174438477,
+      "step": 130
+    },
+    {
+      "epoch": 0.15233949945593037,
+      "grad_norm": 0.03584141284227371,
+      "learning_rate": 0.0001998936986214928,
+      "loss": 0.10438344478607178,
+      "step": 140
+    },
+    {
+      "epoch": 0.1632208922742111,
+      "grad_norm": 0.03158590570092201,
+      "learning_rate": 0.00019983221338380563,
+      "loss": 0.09951521158218384,
+      "step": 150
+    },
+    {
+      "epoch": 0.17410228509249184,
+      "grad_norm": 0.02909119985997677,
+      "learning_rate": 0.0001997567714611106,
+      "loss": 0.09911214709281921,
+      "step": 160
+    },
+    {
+      "epoch": 0.18498367791077258,
+      "grad_norm": 0.030274737626314163,
+      "learning_rate": 0.0001996673834002955,
+      "loss": 0.09334277510643005,
+      "step": 170
+    },
+    {
+      "epoch": 0.19586507072905332,
+      "grad_norm": 0.028922295197844505,
+      "learning_rate": 0.00019956406169793805,
+      "loss": 0.09431872367858887,
+      "step": 180
+    },
+    {
+      "epoch": 0.20674646354733406,
+      "grad_norm": 0.030195271596312523,
+      "learning_rate": 0.00019944682079855867,
+      "loss": 0.10089273452758789,
+      "step": 190
+    },
+    {
+      "epoch": 0.2176278563656148,
+      "grad_norm": 0.029216868802905083,
+      "learning_rate": 0.0001993156770926014,
+      "loss": 0.0949325442314148,
+      "step": 200
+    },
+    {
+      "epoch": 0.2176278563656148,
+      "eval_loss": 0.09785260260105133,
+      "eval_runtime": 5.4576,
+      "eval_samples_per_second": 14.659,
+      "eval_steps_per_second": 1.832,
+      "step": 200
+    },
+    {
+      "epoch": 0.22850924918389554,
+      "grad_norm": 0.029826847836375237,
+      "learning_rate": 0.00019917064891414232,
+      "loss": 0.09331356287002564,
+      "step": 210
+    },
+    {
+      "epoch": 0.23939064200217627,
+      "grad_norm": 0.03143204376101494,
+      "learning_rate": 0.00019901175653832645,
+      "loss": 0.09654195308685302,
+      "step": 220
+    },
+    {
+      "epoch": 0.250272034820457,
+      "grad_norm": 0.02917400933802128,
+      "learning_rate": 0.00019883902217853326,
+      "loss": 0.09114633202552795,
+      "step": 230
+    },
+    {
+      "epoch": 0.2611534276387378,
+      "grad_norm": 0.033463530242443085,
+      "learning_rate": 0.0001986524699832712,
+      "loss": 0.09384369254112243,
+      "step": 240
+    },
+    {
+      "epoch": 0.2720348204570185,
+      "grad_norm": 0.027332467958331108,
+      "learning_rate": 0.00019845212603280174,
+      "loss": 0.08879860639572143,
+      "step": 250
+    },
+    {
+      "epoch": 0.28291621327529926,
+      "grad_norm": 0.03050387091934681,
+      "learning_rate": 0.00019823801833549327,
+      "loss": 0.09065196514129639,
+      "step": 260
+    },
+    {
+      "epoch": 0.29379760609357997,
+      "grad_norm": 0.024866925552487373,
+      "learning_rate": 0.00019801017682390552,
+      "loss": 0.09161882400512696,
+      "step": 270
+    },
+    {
+      "epoch": 0.30467899891186073,
+      "grad_norm": 0.037345923483371735,
+      "learning_rate": 0.0001977686333506049,
+      "loss": 0.09159514904022217,
+      "step": 280
+    },
+    {
+      "epoch": 0.31556039173014144,
+      "grad_norm": 0.0632031187415123,
+      "learning_rate": 0.00019751342168371154,
+      "loss": 0.0902558445930481,
+      "step": 290
+    },
+    {
+      "epoch": 0.3264417845484222,
+      "grad_norm": 0.029686614871025085,
+      "learning_rate": 0.00019724457750217839,
+      "loss": 0.09181964993476868,
+      "step": 300
+    },
+    {
+      "epoch": 0.3373231773667029,
+      "grad_norm": 0.03043133206665516,
+      "learning_rate": 0.0001969621383908033,
+      "loss": 0.08723740577697754,
+      "step": 310
+    },
+    {
+      "epoch": 0.3482045701849837,
+      "grad_norm": 0.041717205196619034,
+      "learning_rate": 0.00019666614383497457,
+      "loss": 0.09164115190505981,
+      "step": 320
+    },
+    {
+      "epoch": 0.3590859630032644,
+      "grad_norm": 0.03025348111987114,
+      "learning_rate": 0.00019635663521515087,
+      "loss": 0.08903414011001587,
+      "step": 330
+    },
+    {
+      "epoch": 0.36996735582154516,
+      "grad_norm": 0.030753403902053833,
+      "learning_rate": 0.00019603365580107615,
+      "loss": 0.08611454963684081,
+      "step": 340
+    },
+    {
+      "epoch": 0.3808487486398259,
+      "grad_norm": 0.021834490820765495,
+      "learning_rate": 0.0001956972507457306,
+      "loss": 0.08958455920219421,
+      "step": 350
+    },
+    {
+      "epoch": 0.39173014145810664,
+      "grad_norm": 0.025724470615386963,
+      "learning_rate": 0.000195347467079018,
+      "loss": 0.08454820513725281,
+      "step": 360
+    },
+    {
+      "epoch": 0.40261153427638735,
+      "grad_norm": 0.034213438630104065,
+      "learning_rate": 0.00019498435370119094,
+      "loss": 0.09067847728729247,
+      "step": 370
+    },
+    {
+      "epoch": 0.4134929270946681,
+      "grad_norm": 0.0437682569026947,
+      "learning_rate": 0.00019460796137601467,
+      "loss": 0.0873767077922821,
+      "step": 380
+    },
+    {
+      "epoch": 0.4243743199129489,
+      "grad_norm": 0.027341296896338463,
+      "learning_rate": 0.00019421834272367,
+      "loss": 0.08544692993164063,
+      "step": 390
+    },
+    {
+      "epoch": 0.4352557127312296,
+      "grad_norm": 0.030272383242845535,
+      "learning_rate": 0.00019381555221339707,
+      "loss": 0.08988077044487,
+      "step": 400
+    },
+    {
+      "epoch": 0.4352557127312296,
+      "eval_loss": 0.09069998562335968,
+      "eval_runtime": 5.3759,
+      "eval_samples_per_second": 14.881,
+      "eval_steps_per_second": 1.86,
+      "step": 400
+    },
+    {
+      "epoch": 0.44613710554951036,
+      "grad_norm": 0.030490044504404068,
+      "learning_rate": 0.00019339964615588043,
+      "loss": 0.088669753074646,
+      "step": 410
+    },
+    {
+      "epoch": 0.45701849836779107,
+      "grad_norm": 0.030257854610681534,
+      "learning_rate": 0.0001929706826953767,
+      "loss": 0.09294685125350952,
+      "step": 420
+    },
+    {
+      "epoch": 0.46789989118607184,
+      "grad_norm": 0.03422554209828377,
+      "learning_rate": 0.00019252872180158605,
+      "loss": 0.08597595095634461,
+      "step": 430
+    },
+    {
+      "epoch": 0.47878128400435255,
+      "grad_norm": 0.026077628135681152,
+      "learning_rate": 0.0001920738252612682,
+      "loss": 0.08706499338150024,
+      "step": 440
+    },
+    {
+      "epoch": 0.4896626768226333,
+      "grad_norm": 0.025868231430649757,
+      "learning_rate": 0.0001916060566696045,
+      "loss": 0.08425920009613037,
+      "step": 450
+    },
+    {
+      "epoch": 0.500544069640914,
+      "grad_norm": 0.03231549635529518,
+      "learning_rate": 0.00019112548142130757,
+      "loss": 0.08545957803726197,
+      "step": 460
+    },
+    {
+      "epoch": 0.5114254624591947,
+      "grad_norm": 0.03557935357093811,
+      "learning_rate": 0.00019063216670147854,
+      "loss": 0.08239468932151794,
+      "step": 470
+    },
+    {
+      "epoch": 0.5223068552774756,
+      "grad_norm": 0.023028500378131866,
+      "learning_rate": 0.00019012618147621488,
+      "loss": 0.08288769721984864,
+      "step": 480
+    },
+    {
+      "epoch": 0.5331882480957563,
+      "grad_norm": 0.021350380033254623,
+      "learning_rate": 0.00018960759648296865,
+      "loss": 0.08916928768157958,
+      "step": 490
+    },
+    {
+      "epoch": 0.544069640914037,
+      "grad_norm": 0.028527546674013138,
+      "learning_rate": 0.0001890764842206573,
+      "loss": 0.09119058847427368,
+      "step": 500
+    },
+    {
+      "epoch": 0.5549510337323177,
+      "grad_norm": 0.023469405248761177,
+      "learning_rate": 0.00018853291893952833,
+      "loss": 0.08246206045150757,
+      "step": 510
+    },
+    {
+      "epoch": 0.5658324265505985,
+      "grad_norm": 0.028494561091065407,
+      "learning_rate": 0.0001879769766307789,
+      "loss": 0.0803530752658844,
+      "step": 520
+    },
+    {
+      "epoch": 0.5767138193688792,
+      "grad_norm": 0.027940815314650536,
+      "learning_rate": 0.00018740873501593214,
+      "loss": 0.08848085403442382,
+      "step": 530
+    },
+    {
+      "epoch": 0.5875952121871599,
+      "grad_norm": 0.02026485837996006,
+      "learning_rate": 0.0001868282735359717,
+      "loss": 0.08710720539093017,
+      "step": 540
+    },
+    {
+      "epoch": 0.5984766050054406,
+      "grad_norm": 0.026024585589766502,
+      "learning_rate": 0.00018623567334023563,
+      "loss": 0.07982800006866456,
+      "step": 550
+    },
+    {
+      "epoch": 0.6093579978237215,
+      "grad_norm": 0.029224518686532974,
+      "learning_rate": 0.0001856310172750717,
+      "loss": 0.08492777943611145,
+      "step": 560
+    },
+    {
+      "epoch": 0.6202393906420022,
+      "grad_norm": 0.02598229981958866,
+      "learning_rate": 0.00018501438987225536,
+      "loss": 0.08290873765945435,
+      "step": 570
+    },
+    {
+      "epoch": 0.6311207834602829,
+      "grad_norm": 0.023201750591397285,
+      "learning_rate": 0.00018438587733717187,
+      "loss": 0.08711419701576233,
+      "step": 580
+    },
+    {
+      "epoch": 0.6420021762785637,
+      "grad_norm": 0.026234455406665802,
+      "learning_rate": 0.00018374556753676506,
+      "loss": 0.08342958688735962,
+      "step": 590
+    },
+    {
+      "epoch": 0.6528835690968444,
+      "grad_norm": 0.021891364827752113,
+      "learning_rate": 0.00018309354998725306,
+      "loss": 0.07860854864120484,
+      "step": 600
+    },
+    {
+      "epoch": 0.6528835690968444,
+      "eval_loss": 0.08754301071166992,
+      "eval_runtime": 5.5341,
+      "eval_samples_per_second": 14.456,
+      "eval_steps_per_second": 1.807,
+      "step": 600
+    },
+    {
+      "epoch": 0.6637649619151251,
+      "grad_norm": 0.026508808135986328,
+      "learning_rate": 0.00018242991584161393,
+      "loss": 0.08734351992607117,
+      "step": 610
+    },
+    {
+      "epoch": 0.6746463547334058,
+      "grad_norm": 0.02286313846707344,
+      "learning_rate": 0.0001817547578768424,
+      "loss": 0.08437882661819458,
+      "step": 620
+    },
+    {
+      "epoch": 0.6855277475516867,
+      "grad_norm": 0.02370450273156166,
+      "learning_rate": 0.00018106817048097937,
+      "loss": 0.08016583323478699,
+      "step": 630
+    },
+    {
+      "epoch": 0.6964091403699674,
+      "grad_norm": 0.025123964995145798,
+      "learning_rate": 0.0001803702496399165,
+      "loss": 0.076961088180542,
+      "step": 640
+    },
+    {
+      "epoch": 0.7072905331882481,
+      "grad_norm": 0.019611909985542297,
+      "learning_rate": 0.0001796610929239771,
+      "loss": 0.08545655012130737,
+      "step": 650
+    },
+    {
+      "epoch": 0.7181719260065288,
+      "grad_norm": 0.020668957382440567,
+      "learning_rate": 0.00017894079947427573,
+      "loss": 0.0809054434299469,
+      "step": 660
+    },
+    {
+      "epoch": 0.7290533188248096,
+      "grad_norm": 0.028357358649373055,
+      "learning_rate": 0.0001782094699888582,
+      "loss": 0.07935836911201477,
+      "step": 670
+    },
+    {
+      "epoch": 0.7399347116430903,
+      "grad_norm": 0.018039165064692497,
+      "learning_rate": 0.00017746720670862363,
+      "loss": 0.08474135994911194,
+      "step": 680
+    },
+    {
+      "epoch": 0.750816104461371,
+      "grad_norm": 0.02181115373969078,
+      "learning_rate": 0.0001767141134030313,
+      "loss": 0.08088613748550415,
+      "step": 690
+    },
+    {
+      "epoch": 0.7616974972796517,
+      "grad_norm": 0.029043374583125114,
+      "learning_rate": 0.00017595029535559336,
+      "loss": 0.078003191947937,
+      "step": 700
+    },
+    {
+      "epoch": 0.7725788900979326,
+      "grad_norm": 0.024250885471701622,
+      "learning_rate": 0.00017517585934915595,
+      "loss": 0.08291269540786743,
+      "step": 710
+    },
+    {
+      "epoch": 0.7834602829162133,
+      "grad_norm": 0.028013000264763832,
+      "learning_rate": 0.00017439091365097106,
+      "loss": 0.08375194072723388,
+      "step": 720
+    },
+    {
+      "epoch": 0.794341675734494,
+      "grad_norm": 0.030651913955807686,
+      "learning_rate": 0.00017359556799756052,
+      "loss": 0.08069210052490235,
+      "step": 730
+    },
+    {
+      "epoch": 0.8052230685527747,
+      "grad_norm": 0.028212785720825195,
+      "learning_rate": 0.00017278993357937455,
+      "loss": 0.08342626094818115,
+      "step": 740
+    },
+    {
+      "epoch": 0.8161044613710555,
+      "grad_norm": 0.024551352486014366,
+      "learning_rate": 0.0001719741230252473,
+      "loss": 0.0814591646194458,
+      "step": 750
+    },
+    {
+      "epoch": 0.8269858541893362,
+      "grad_norm": 0.022670986130833626,
+      "learning_rate": 0.00017114825038665123,
+      "loss": 0.08308478593826293,
+      "step": 760
+    },
+    {
+      "epoch": 0.8378672470076169,
+      "grad_norm": 0.03288818150758743,
+      "learning_rate": 0.0001703124311217524,
+      "loss": 0.07733856439590454,
+      "step": 770
+    },
+    {
+      "epoch": 0.8487486398258978,
+      "grad_norm": 0.02053298056125641,
+      "learning_rate": 0.00016946678207926938,
+      "loss": 0.0828587532043457,
+      "step": 780
+    },
+    {
+      "epoch": 0.8596300326441785,
+      "grad_norm": 0.024455999955534935,
+      "learning_rate": 0.0001686114214821375,
+      "loss": 0.07782068252563476,
+      "step": 790
+    },
+    {
+      "epoch": 0.8705114254624592,
+      "grad_norm": 0.022575655952095985,
+      "learning_rate": 0.00016774646891098134,
+      "loss": 0.08273504972457886,
+      "step": 800
+    },
+    {
+      "epoch": 0.8705114254624592,
+      "eval_loss": 0.08349094539880753,
+      "eval_runtime": 5.5064,
+      "eval_samples_per_second": 14.529,
+      "eval_steps_per_second": 1.816,
+      "step": 800
+    },
+    {
+      "epoch": 0.8813928182807399,
+      "grad_norm": 0.02310466580092907,
+      "learning_rate": 0.00016687204528739694,
+      "loss": 0.08457955718040466,
+      "step": 810
+    },
+    {
+      "epoch": 0.8922742110990207,
+      "grad_norm": 0.02504105120897293,
+      "learning_rate": 0.00016598827285704696,
+      "loss": 0.08319224119186401,
+      "step": 820
+    },
+    {
+      "epoch": 0.9031556039173014,
+      "grad_norm": 0.02492046356201172,
+      "learning_rate": 0.00016509527517257042,
+      "loss": 0.08071508407592773,
+      "step": 830
+    },
+    {
+      "epoch": 0.9140369967355821,
+      "grad_norm": 0.023182213306427002,
+      "learning_rate": 0.0001641931770763101,
+      "loss": 0.08242483139038086,
+      "step": 840
+    },
+    {
+      "epoch": 0.9249183895538629,
+      "grad_norm": 0.022171733900904655,
+      "learning_rate": 0.00016328210468285911,
+      "loss": 0.08000438213348389,
+      "step": 850
+    },
+    {
+      "epoch": 0.9357997823721437,
+      "grad_norm": 0.023160789161920547,
+      "learning_rate": 0.00016236218536143013,
+      "loss": 0.08225069642066955,
+      "step": 860
+    },
+    {
+      "epoch": 0.9466811751904244,
+      "grad_norm": 0.021988507360219955,
+      "learning_rate": 0.0001614335477180488,
+      "loss": 0.07968268394470215,
+      "step": 870
+    },
+    {
+      "epoch": 0.9575625680087051,
+      "grad_norm": 0.0209442637860775,
+      "learning_rate": 0.00016049632157757473,
+      "loss": 0.08057799339294433,
+      "step": 880
+    },
+    {
+      "epoch": 0.9684439608269858,
+      "grad_norm": 0.02684466354548931,
+      "learning_rate": 0.00015955063796555137,
+      "loss": 0.07912254929542542,
+      "step": 890
+    },
+    {
+      "epoch": 0.9793253536452666,
+      "grad_norm": 0.022005856037139893,
+      "learning_rate": 0.00015859662908988885,
+      "loss": 0.07660449743270874,
+      "step": 900
+    },
+    {
+      "epoch": 0.9902067464635473,
+      "grad_norm": 0.022977208718657494,
+      "learning_rate": 0.0001576344283223809,
+      "loss": 0.07685725092887878,
+      "step": 910
+    },
+    {
+      "epoch": 1.001088139281828,
+      "grad_norm": 0.020924242213368416,
+      "learning_rate": 0.00015666417018005938,
+      "loss": 0.07933401465415954,
+      "step": 920
+    },
+    {
+      "epoch": 1.0119695321001088,
+      "grad_norm": 0.02789444290101528,
+      "learning_rate": 0.00015568599030638854,
+      "loss": 0.07339736819267273,
+      "step": 930
+    },
+    {
+      "epoch": 1.0228509249183895,
+      "grad_norm": 0.02638629637658596,
+      "learning_rate": 0.00015470002545230193,
+      "loss": 0.07634537220001221,
+      "step": 940
+    },
+    {
+      "epoch": 1.0337323177366704,
+      "grad_norm": 0.029259871691465378,
+      "learning_rate": 0.0001537064134570844,
+      "loss": 0.08004393577575683,
+      "step": 950
+    },
+    {
+      "epoch": 1.0446137105549511,
+      "grad_norm": 0.03044624626636505,
+      "learning_rate": 0.0001527052932291019,
+      "loss": 0.07677503824234008,
+      "step": 960
+    },
+    {
+      "epoch": 1.0554951033732318,
+      "grad_norm": 0.02535896934568882,
+      "learning_rate": 0.00015169680472638218,
+      "loss": 0.07567052841186524,
+      "step": 970
+    },
+    {
+      "epoch": 1.0663764961915125,
+      "grad_norm": 0.02594912238419056,
+      "learning_rate": 0.00015068108893704797,
+      "loss": 0.0768008291721344,
+      "step": 980
+    },
+    {
+      "epoch": 1.0772578890097932,
+      "grad_norm": 0.026760386303067207,
+      "learning_rate": 0.0001496582878596071,
+      "loss": 0.07604062557220459,
+      "step": 990
+    },
+    {
+      "epoch": 1.088139281828074,
+      "grad_norm": 0.02427523396909237,
+      "learning_rate": 0.00014862854448310074,
+      "loss": 0.07631718516349792,
+      "step": 1000
+    },
+    {
+      "epoch": 1.088139281828074,
+      "eval_loss": 0.08224387466907501,
+      "eval_runtime": 5.3767,
+      "eval_samples_per_second": 14.879,
+      "eval_steps_per_second": 1.86,
+      "step": 1000
+    },
+    {
+      "epoch": 1.0990206746463547,
+      "grad_norm": 0.023537471890449524,
+      "learning_rate": 0.00014759200276711316,
+      "loss": 0.0752741813659668,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1099020674646354,
+      "grad_norm": 0.028346385806798935,
+      "learning_rate": 0.00014654880762164624,
+      "loss": 0.07320383191108704,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1207834602829163,
+      "grad_norm": 0.024628447368741035,
+      "learning_rate": 0.00014549910488686078,
+      "loss": 0.08079155683517455,
+      "step": 1030
+    },
+    {
+      "epoch": 1.131664853101197,
+      "grad_norm": 0.024834809824824333,
+      "learning_rate": 0.0001444430413126877,
+      "loss": 0.07409957647323609,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1425462459194777,
+      "grad_norm": 0.03418092057108879,
+      "learning_rate": 0.0001433807645383123,
+      "loss": 0.07469584345817566,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1534276387377584,
+      "grad_norm": 0.022386226803064346,
+      "learning_rate": 0.00014231242307153417,
+      "loss": 0.07519133687019348,
+      "step": 1060
+    },
+    {
+      "epoch": 1.1643090315560392,
+      "grad_norm": 0.023596055805683136,
+      "learning_rate": 0.00014123816626800535,
+      "loss": 0.07806066274642945,
+      "step": 1070
+    },
+    {
+      "epoch": 1.1751904243743199,
+      "grad_norm": 0.025627054274082184,
+      "learning_rate": 0.00014015814431035046,
+      "loss": 0.07402833104133606,
+      "step": 1080
+    },
+    {
+      "epoch": 1.1860718171926006,
+      "grad_norm": 0.021048523485660553,
+      "learning_rate": 0.00013907250818717082,
+      "loss": 0.07758893370628357,
+      "step": 1090
+    },
+    {
+      "epoch": 1.1969532100108813,
+      "grad_norm": 0.028320247307419777,
+      "learning_rate": 0.00013798140967193607,
+      "loss": 0.0765085220336914,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2078346028291622,
+      "grad_norm": 0.027219558134675026,
+      "learning_rate": 0.0001368850013017659,
+      "loss": 0.07446168661117554,
+      "step": 1110
+    },
+    {
+      "epoch": 1.218715995647443,
+      "grad_norm": 0.022490857169032097,
+      "learning_rate": 0.0001357834363561054,
+      "loss": 0.07697376012802123,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2295973884657236,
+      "grad_norm": 0.02330136112868786,
+      "learning_rate": 0.00013467686883529616,
+      "loss": 0.07688432335853576,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2404787812840044,
+      "grad_norm": 0.025878841057419777,
+      "learning_rate": 0.0001335654534390468,
+      "loss": 0.07023751735687256,
+      "step": 1140
+    },
+    {
+      "epoch": 1.251360174102285,
+      "grad_norm": 0.023564936593174934,
+      "learning_rate": 0.0001324493455448059,
+      "loss": 0.0761339783668518,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2622415669205658,
+      "grad_norm": 0.01974901743233204,
+      "learning_rate": 0.0001313287011860399,
+      "loss": 0.07061100006103516,
+      "step": 1160
+    },
+    {
+      "epoch": 1.2731229597388465,
+      "grad_norm": 0.028187213465571404,
+      "learning_rate": 0.00013020367703041934,
+      "loss": 0.07600026726722717,
+      "step": 1170
+    },
+    {
+      "epoch": 1.2840043525571274,
+      "grad_norm": 0.027759388089179993,
+      "learning_rate": 0.0001290744303579167,
+      "loss": 0.07545133829116821,
+      "step": 1180
+    },
+    {
+      "epoch": 1.2948857453754081,
+      "grad_norm": 0.03569910302758217,
+      "learning_rate": 0.00012794111903881818,
+      "loss": 0.07300611138343811,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3057671381936888,
+      "grad_norm": 0.026020081713795662,
+      "learning_rate": 0.00012680390151165337,
+      "loss": 0.07233045101165772,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3057671381936888,
+      "eval_loss": 0.08009672164916992,
+      "eval_runtime": 5.448,
+      "eval_samples_per_second": 14.684,
+      "eval_steps_per_second": 1.836,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3166485310119695,
+      "grad_norm": 0.02722887694835663,
+      "learning_rate": 0.0001256629367610454,
+      "loss": 0.0722274899482727,
+      "step": 1210
+    },
+    {
+      "epoch": 1.3275299238302503,
+      "grad_norm": 0.021629413589835167,
+      "learning_rate": 0.0001245183842954844,
+      "loss": 0.07423741221427918,
+      "step": 1220
+    },
+    {
+      "epoch": 1.338411316648531,
+      "grad_norm": 0.030425263568758965,
+      "learning_rate": 0.00012337040412502823,
+      "loss": 0.07771967649459839,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3492927094668117,
+      "grad_norm": 0.022577406838536263,
+      "learning_rate": 0.00012221915673893267,
+      "loss": 0.07685770392417908,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3601741022850926,
+      "grad_norm": 0.026748720556497574,
+      "learning_rate": 0.00012106480308321498,
+      "loss": 0.0744680941104889,
+      "step": 1250
+    },
+    {
+      "epoch": 1.371055495103373,
+      "grad_norm": 0.024764206260442734,
+      "learning_rate": 0.00011990750453815324,
+      "loss": 0.07823204398155212,
+      "step": 1260
+    },
+    {
+      "epoch": 1.381936887921654,
+      "grad_norm": 0.023767156526446342,
+      "learning_rate": 0.00011874742289572531,
+      "loss": 0.06924179792404175,
+      "step": 1270
+    },
+    {
+      "epoch": 1.3928182807399347,
+      "grad_norm": 0.025964556261897087,
+      "learning_rate": 0.00011758472033698993,
+      "loss": 0.06726616621017456,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4036996735582155,
+      "grad_norm": 0.024118445813655853,
+      "learning_rate": 0.00011641955940941374,
+      "loss": 0.07383161187171935,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4145810663764962,
+      "grad_norm": 0.0229018721729517,
+      "learning_rate": 0.00011525210300414673,
+      "loss": 0.07341222763061524,
+      "step": 1300
+    },
+    {
+      "epoch": 1.4254624591947769,
+      "grad_norm": 0.02409880980849266,
+      "learning_rate": 0.00011408251433325006,
+      "loss": 0.07354844808578491,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4363438520130578,
+      "grad_norm": 0.039964865893125534,
+      "learning_rate": 0.00011291095690687847,
+      "loss": 0.07191305160522461,
+      "step": 1320
+    },
+    {
+      "epoch": 1.4472252448313383,
+      "grad_norm": 0.02640443667769432,
+      "learning_rate": 0.00011173759451042168,
+      "loss": 0.07049686908721924,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4581066376496192,
+      "grad_norm": 0.022137323394417763,
+      "learning_rate": 0.0001105625911816067,
+      "loss": 0.07753136157989501,
+      "step": 1340
+    },
+    {
+      "epoch": 1.4689880304679,
+      "grad_norm": 0.02352975122630596,
+      "learning_rate": 0.0001093861111875652,
+      "loss": 0.07463480234146118,
+      "step": 1350
+    },
+    {
+      "epoch": 1.4798694232861807,
+      "grad_norm": 0.02214525081217289,
+      "learning_rate": 0.0001082083190018688,
+      "loss": 0.06845179796218873,
+      "step": 1360
+    },
+    {
+      "epoch": 1.4907508161044614,
+      "grad_norm": 0.025555888190865517,
+      "learning_rate": 0.00010702937928153533,
+      "loss": 0.07124170064926147,
+      "step": 1370
+    },
+    {
+      "epoch": 1.501632208922742,
+      "grad_norm": 0.02631489560008049,
+      "learning_rate": 0.00010584945684400963,
+      "loss": 0.07157056331634522,
+      "step": 1380
+    },
+    {
+      "epoch": 1.512513601741023,
+      "grad_norm": 0.026430578902363777,
+      "learning_rate": 0.0001046687166441217,
+      "loss": 0.07220081686973571,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5233949945593035,
+      "grad_norm": 0.024772843345999718,
+      "learning_rate": 0.00010348732375102601,
+      "loss": 0.07243863940238952,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5233949945593035,
+      "eval_loss": 0.07733941823244095,
+      "eval_runtime": 5.3743,
+      "eval_samples_per_second": 14.886,
+      "eval_steps_per_second": 1.861,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5342763873775844,
+      "grad_norm": 0.0237529706209898,
+      "learning_rate": 0.00010230544332512442,
+      "loss": 0.06972206830978393,
+      "step": 1410
+    },
+    {
+      "epoch": 1.545157780195865,
+      "grad_norm": 0.02250276878476143,
+      "learning_rate": 0.00010112324059497649,
+      "loss": 0.07575106620788574,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5560391730141458,
+      "grad_norm": 0.02346464991569519,
+      "learning_rate": 9.994088083420044e-05,
+      "loss": 0.07188562154769898,
+      "step": 1430
+    },
+    {
+      "epoch": 1.5669205658324266,
+      "grad_norm": 0.023935239762067795,
+      "learning_rate": 9.875852933836765e-05,
+      "loss": 0.0751798152923584,
+      "step": 1440
+    },
+    {
+      "epoch": 1.5778019586507073,
+      "grad_norm": 0.02204659767448902,
+      "learning_rate": 9.757635140189394e-05,
+      "loss": 0.06870698928833008,
+      "step": 1450
+    },
+    {
+      "epoch": 1.588683351468988,
+      "grad_norm": 0.022586360573768616,
+      "learning_rate": 9.639451229493133e-05,
+      "loss": 0.07310816049575805,
+      "step": 1460
+    },
+    {
+      "epoch": 1.5995647442872687,
+      "grad_norm": 0.02241821214556694,
+      "learning_rate": 9.521317724026299e-05,
+      "loss": 0.07275789976119995,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6104461371055496,
+      "grad_norm": 0.019981661811470985,
+      "learning_rate": 9.403251139020486e-05,
+      "loss": 0.07513962984085083,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6213275299238301,
+      "grad_norm": 0.0203393641859293,
+      "learning_rate": 9.285267980351719e-05,
+      "loss": 0.0734050989151001,
+      "step": 1490
+    },
+    {
+      "epoch": 1.632208922742111,
+      "grad_norm": 0.027680471539497375,
+      "learning_rate": 9.167384742232898e-05,
+      "loss": 0.06607317924499512,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6430903155603918,
+      "grad_norm": 0.026372363790869713,
+      "learning_rate": 9.049617904907877e-05,
+      "loss": 0.07146948575973511,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6539717083786725,
+      "grad_norm": 0.02281905524432659,
+      "learning_rate": 8.931983932347533e-05,
+      "loss": 0.07334067225456238,
+      "step": 1520
+    },
+    {
+      "epoch": 1.6648531011969532,
+      "grad_norm": 0.02359774336218834,
+      "learning_rate": 8.814499269948053e-05,
+      "loss": 0.0697287678718567,
+      "step": 1530
+    },
+    {
+      "epoch": 1.6757344940152339,
+      "grad_norm": 0.02269565500319004,
+      "learning_rate": 8.697180342231849e-05,
+      "loss": 0.07186954021453858,
+      "step": 1540
+    },
+    {
+      "epoch": 1.6866158868335148,
+      "grad_norm": 0.028245631605386734,
+      "learning_rate": 8.580043550551393e-05,
+      "loss": 0.06877661943435669,
+      "step": 1550
+    },
+    {
+      "epoch": 1.6974972796517953,
+      "grad_norm": 0.02352772280573845,
+      "learning_rate": 8.463105270796276e-05,
+      "loss": 0.07183843851089478,
+      "step": 1560
+    },
+    {
+      "epoch": 1.7083786724700762,
+      "grad_norm": 0.023996759206056595,
+      "learning_rate": 8.346381851103842e-05,
+      "loss": 0.07082110047340393,
+      "step": 1570
+    },
+    {
+      "epoch": 1.719260065288357,
+      "grad_norm": 0.025422753766179085,
+      "learning_rate": 8.229889609573699e-05,
+      "loss": 0.06895624399185181,
+      "step": 1580
+    },
+    {
+      "epoch": 1.7301414581066377,
+      "grad_norm": 0.02499948814511299,
+      "learning_rate": 8.113644831986416e-05,
+      "loss": 0.0707672655582428,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7410228509249184,
+      "grad_norm": 0.030967354774475098,
+      "learning_rate": 7.997663769526753e-05,
+      "loss": 0.07331972122192383,
+      "step": 1600
+    },
+    {
+      "epoch": 1.7410228509249184,
+      "eval_loss": 0.07515527307987213,
+      "eval_runtime": 5.3812,
+      "eval_samples_per_second": 14.867,
+      "eval_steps_per_second": 1.858,
+      "step": 1600
+    },
+    {
+      "epoch": 1.751904243743199,
+      "grad_norm": 0.026022454723715782,
+      "learning_rate": 7.881962636511715e-05,
+      "loss": 0.07261155843734741,
+      "step": 1610
+    },
+    {
+      "epoch": 1.76278563656148,
+      "grad_norm": 0.03104907087981701,
+      "learning_rate": 7.766557608123783e-05,
+      "loss": 0.07076815962791443,
+      "step": 1620
+    },
+    {
+      "epoch": 1.7736670293797605,
+      "grad_norm": 0.026025639846920967,
+      "learning_rate": 7.651464818149587e-05,
+      "loss": 0.0643875777721405,
+      "step": 1630
+    },
+    {
+      "epoch": 1.7845484221980414,
+      "grad_norm": 0.022654540836811066,
+      "learning_rate": 7.536700356724379e-05,
+      "loss": 0.07072726488113404,
+      "step": 1640
+    },
+    {
+      "epoch": 1.795429815016322,
+      "grad_norm": 0.02240171656012535,
+      "learning_rate": 7.422280268082629e-05,
+      "loss": 0.07485873699188232,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8063112078346029,
+      "grad_norm": 0.02509446069598198,
+      "learning_rate": 7.308220548314981e-05,
+      "loss": 0.06636062860488892,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8171926006528836,
+      "grad_norm": 0.023968016728758812,
+      "learning_rate": 7.194537143132023e-05,
+      "loss": 0.06702766418457032,
+      "step": 1670
+    },
+    {
+      "epoch": 1.8280739934711643,
+      "grad_norm": 0.027016611769795418,
+      "learning_rate": 7.081245945635013e-05,
+      "loss": 0.07087734937667847,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8389553862894452,
+      "grad_norm": 0.02593705803155899,
+      "learning_rate": 6.968362794094027e-05,
+      "loss": 0.06891222596168518,
+      "step": 1690
+    },
+    {
+      "epoch": 1.8498367791077257,
+      "grad_norm": 0.02612975798547268,
+      "learning_rate": 6.855903469733738e-05,
+      "loss": 0.06720860004425049,
+      "step": 1700
+    },
+    {
+      "epoch": 1.8607181719260066,
+      "grad_norm": 0.025018637999892235,
+      "learning_rate": 6.743883694527169e-05,
+      "loss": 0.06776891946792603,
+      "step": 1710
+    },
+    {
+      "epoch": 1.8715995647442871,
+      "grad_norm": 0.02438053861260414,
+      "learning_rate": 6.632319128997759e-05,
+      "loss": 0.07007836103439331,
+      "step": 1720
+    },
+    {
+      "epoch": 1.882480957562568,
+      "grad_norm": 0.02692045457661152,
+      "learning_rate": 6.52122537002998e-05,
+      "loss": 0.07095590829849244,
+      "step": 1730
+    },
+    {
+      "epoch": 1.8933623503808488,
+      "grad_norm": 0.031177183613181114,
+      "learning_rate": 6.410617948688892e-05,
+      "loss": 0.06855381727218628,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9042437431991295,
+      "grad_norm": 0.022553348913788795,
+      "learning_rate": 6.300512328048845e-05,
+      "loss": 0.07089903354644775,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9151251360174102,
+      "grad_norm": 0.024013947695493698,
+      "learning_rate": 6.190923901031757e-05,
+      "loss": 0.0708732008934021,
+      "step": 1760
+    },
+    {
+      "epoch": 1.926006528835691,
+      "grad_norm": 0.022148922085762024,
+      "learning_rate": 6.081867988255135e-05,
+      "loss": 0.07153573036193847,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9368879216539718,
+      "grad_norm": 0.047587718814611435,
+      "learning_rate": 5.973359835890246e-05,
+      "loss": 0.0697596788406372,
+      "step": 1780
+    },
+    {
+      "epoch": 1.9477693144722523,
+      "grad_norm": 0.02610265091061592,
+      "learning_rate": 5.865414613530669e-05,
+      "loss": 0.06653320789337158,
+      "step": 1790
+    },
+    {
+      "epoch": 1.9586507072905333,
+      "grad_norm": 0.02527604252099991,
+      "learning_rate": 5.758047412071568e-05,
+      "loss": 0.07076951861381531,
+      "step": 1800
+    },
+    {
+      "epoch": 1.9586507072905333,
+      "eval_loss": 0.07202806323766708,
+      "eval_runtime": 5.4058,
+      "eval_samples_per_second": 14.799,
+      "eval_steps_per_second": 1.85,
+      "step": 1800
+    },
+    {
+      "epoch": 1.969532100108814,
+      "grad_norm": 0.024923427030444145,
+      "learning_rate": 5.651273241599976e-05,
+      "loss": 0.06910277605056762,
+      "step": 1810
+    },
+    {
+      "epoch": 1.9804134929270947,
+      "grad_norm": 0.023715443909168243,
+      "learning_rate": 5.5451070292963434e-05,
+      "loss": 0.06721980571746826,
+      "step": 1820
+    },
+    {
+      "epoch": 1.9912948857453754,
+      "grad_norm": 0.019582392647862434,
+      "learning_rate": 5.439563617347711e-05,
+      "loss": 0.06764072179794312,
+      "step": 1830
+    },
+    {
+      "epoch": 2.002176278563656,
+      "grad_norm": 0.027950329706072807,
+      "learning_rate": 5.3346577608727463e-05,
+      "loss": 0.07124176025390624,
+      "step": 1840
+    },
+    {
+      "epoch": 2.013057671381937,
+      "grad_norm": 0.036231789737939835,
+      "learning_rate": 5.230404125858952e-05,
+      "loss": 0.05957789421081543,
+      "step": 1850
+    },
+    {
+      "epoch": 2.0239390642002175,
+      "grad_norm": 0.026198675855994225,
+      "learning_rate": 5.1268172871123535e-05,
+      "loss": 0.055424296855926515,
+      "step": 1860
+    },
+    {
+      "epoch": 2.0348204570184985,
+      "grad_norm": 0.025477459654211998,
+      "learning_rate": 5.023911726219902e-05,
+      "loss": 0.06114422678947449,
+      "step": 1870
+    },
+    {
+      "epoch": 2.045701849836779,
+      "grad_norm": 0.03552868962287903,
+      "learning_rate": 4.921701829524956e-05,
+      "loss": 0.058832204341888426,
+      "step": 1880
+    },
+    {
+      "epoch": 2.05658324265506,
+      "grad_norm": 0.03449210524559021,
+      "learning_rate": 4.820201886116022e-05,
+      "loss": 0.057713437080383304,
+      "step": 1890
+    },
+    {
+      "epoch": 2.067464635473341,
+      "grad_norm": 0.033065732568502426,
+      "learning_rate": 4.719426085829123e-05,
+      "loss": 0.05968807935714722,
+      "step": 1900
+    },
+    {
+      "epoch": 2.0783460282916213,
+      "grad_norm": 0.03002762608230114,
+      "learning_rate": 4.619388517264056e-05,
+      "loss": 0.060094785690307614,
+      "step": 1910
+    },
+    {
+      "epoch": 2.0892274211099022,
+      "grad_norm": 0.034517038613557816,
+      "learning_rate": 4.5201031658147766e-05,
+      "loss": 0.05601954460144043,
+      "step": 1920
+    },
+    {
+      "epoch": 2.1001088139281827,
+      "grad_norm": 0.03293369337916374,
+      "learning_rate": 4.4215839117142134e-05,
+      "loss": 0.06034954190254212,
+      "step": 1930
+    },
+    {
+      "epoch": 2.1109902067464636,
+      "grad_norm": 0.032126300036907196,
+      "learning_rate": 4.3238445280938154e-05,
+      "loss": 0.05510505437850952,
+      "step": 1940
+    },
+    {
+      "epoch": 2.121871599564744,
+      "grad_norm": 0.027177443727850914,
+      "learning_rate": 4.2268986790580125e-05,
+      "loss": 0.060755813121795656,
+      "step": 1950
+    },
+    {
+      "epoch": 2.132752992383025,
+      "grad_norm": 0.030124740675091743,
+      "learning_rate": 4.130759917773985e-05,
+      "loss": 0.06080647706985474,
+      "step": 1960
+    },
+    {
+      "epoch": 2.143634385201306,
+      "grad_norm": 0.03235238790512085,
+      "learning_rate": 4.0354416845769e-05,
+      "loss": 0.06485359072685241,
+      "step": 1970
+    },
+    {
+      "epoch": 2.1545157780195865,
+      "grad_norm": 0.033107586205005646,
+      "learning_rate": 3.940957305090935e-05,
+      "loss": 0.059920203685760495,
+      "step": 1980
+    },
+    {
+      "epoch": 2.1653971708378674,
+      "grad_norm": 0.02939060889184475,
+      "learning_rate": 3.8473199883663334e-05,
+      "loss": 0.0490723729133606,
+      "step": 1990
+    },
+    {
+      "epoch": 2.176278563656148,
+      "grad_norm": 0.03297962248325348,
+      "learning_rate": 3.754542825032766e-05,
+      "loss": 0.053455913066864015,
+      "step": 2000
+    },
+    {
+      "epoch": 2.176278563656148,
+      "eval_loss": 0.07033731788396835,
+      "eval_runtime": 5.4006,
+      "eval_samples_per_second": 14.813,
+      "eval_steps_per_second": 1.852,
+      "step": 2000
+    },
+    {
+      "epoch": 2.187159956474429,
+      "grad_norm": 0.026707574725151062,
+      "learning_rate": 3.6626387854692365e-05,
+      "loss": 0.061966001987457275,
+      "step": 2010
+    },
+    {
+      "epoch": 2.1980413492927093,
+      "grad_norm": 0.03455955162644386,
+      "learning_rate": 3.571620717990811e-05,
+      "loss": 0.05781112909317017,
+      "step": 2020
+    },
+    {
+      "epoch": 2.2089227421109903,
+      "grad_norm": 0.02576427161693573,
+      "learning_rate": 3.4815013470524004e-05,
+      "loss": 0.060006117820739745,
+      "step": 2030
+    },
+    {
+      "epoch": 2.2198041349292708,
+      "grad_norm": 0.03262777253985405,
+      "learning_rate": 3.392293271469866e-05,
+      "loss": 0.05959203839302063,
+      "step": 2040
+    },
+    {
+      "epoch": 2.2306855277475517,
+      "grad_norm": 0.031093895435333252,
+      "learning_rate": 3.304008962658677e-05,
+      "loss": 0.057662928104400636,
+      "step": 2050
+    },
+    {
+      "epoch": 2.2415669205658326,
+      "grad_norm": 0.030908960849046707,
+      "learning_rate": 3.216660762890413e-05,
+      "loss": 0.05794321894645691,
+      "step": 2060
+    },
+    {
+      "epoch": 2.252448313384113,
+      "grad_norm": 0.03053661808371544,
+      "learning_rate": 3.130260883567272e-05,
+      "loss": 0.05402252674102783,
+      "step": 2070
+    },
+    {
+      "epoch": 2.263329706202394,
+      "grad_norm": 0.03832336515188217,
+      "learning_rate": 3.0448214035149102e-05,
+      "loss": 0.06356992721557617,
+      "step": 2080
+    },
+    {
+      "epoch": 2.2742110990206745,
+      "grad_norm": 0.035567495971918106,
+      "learning_rate": 2.960354267293809e-05,
+      "loss": 0.057219624519348145,
+      "step": 2090
+    },
+    {
+      "epoch": 2.2850924918389555,
+      "grad_norm": 0.03574289754033089,
+      "learning_rate": 2.8768712835293922e-05,
+      "loss": 0.0654039740562439,
+      "step": 2100
+    },
+    {
+      "epoch": 2.295973884657236,
+      "grad_norm": 0.03358747810125351,
+      "learning_rate": 2.7943841232611812e-05,
+      "loss": 0.05935961604118347,
+      "step": 2110
+    },
+    {
+      "epoch": 2.306855277475517,
+      "grad_norm": 0.036444805562496185,
+      "learning_rate": 2.712904318311167e-05,
+      "loss": 0.059474557638168335,
+      "step": 2120
+    },
+    {
+      "epoch": 2.3177366702937974,
+      "grad_norm": 0.04165448620915413,
+      "learning_rate": 2.6324432596716276e-05,
+      "loss": 0.06448624134063721,
+      "step": 2130
+    },
+    {
+      "epoch": 2.3286180631120783,
+      "grad_norm": 0.03665981814265251,
+      "learning_rate": 2.5530121959126718e-05,
+      "loss": 0.055650746822357176,
+      "step": 2140
+    },
+    {
+      "epoch": 2.3394994559303592,
+      "grad_norm": 0.02873632311820984,
+      "learning_rate": 2.47462223160965e-05,
+      "loss": 0.05966170430183411,
+      "step": 2150
+    },
+    {
+      "epoch": 2.3503808487486397,
+      "grad_norm": 0.03279171139001846,
+      "learning_rate": 2.397284325790744e-05,
+      "loss": 0.06053815484046936,
+      "step": 2160
+    },
+    {
+      "epoch": 2.3612622415669207,
+      "grad_norm": 0.03719950094819069,
+      "learning_rate": 2.321009290404864e-05,
+      "loss": 0.0560091495513916,
+      "step": 2170
+    },
+    {
+      "epoch": 2.372143634385201,
+      "grad_norm": 0.03888888284564018,
+      "learning_rate": 2.2458077888101226e-05,
+      "loss": 0.058753132820129395,
+      "step": 2180
+    },
+    {
+      "epoch": 2.383025027203482,
+      "grad_norm": 0.0321931354701519,
+      "learning_rate": 2.1716903342830995e-05,
+      "loss": 0.05801019072532654,
+      "step": 2190
+    },
+    {
+      "epoch": 2.3939064200217626,
+      "grad_norm": 0.038627929985523224,
+      "learning_rate": 2.0986672885490378e-05,
+      "loss": 0.06117302179336548,
+      "step": 2200
+    },
+    {
+      "epoch": 2.3939064200217626,
+      "eval_loss": 0.0698942095041275,
+      "eval_runtime": 5.4996,
+      "eval_samples_per_second": 14.546,
+      "eval_steps_per_second": 1.818,
+      "step": 2200
+    },
+    {
+      "epoch": 2.4047878128400435,
+      "grad_norm": 0.03631049767136574,
+      "learning_rate": 2.0267488603332907e-05,
+      "loss": 0.05466359853744507,
+      "step": 2210
+    },
+    {
+      "epoch": 2.4156692056583244,
+      "grad_norm": 0.03237799555063248,
+      "learning_rate": 1.9559451039341147e-05,
+      "loss": 0.05731322765350342,
+      "step": 2220
+    },
+    {
+      "epoch": 2.426550598476605,
+      "grad_norm": 0.0349552221596241,
+      "learning_rate": 1.886265917817064e-05,
+      "loss": 0.06014159917831421,
+      "step": 2230
+    },
+    {
+      "epoch": 2.437431991294886,
+      "grad_norm": 0.0360136479139328,
+      "learning_rate": 1.8177210432311732e-05,
+      "loss": 0.05688718557357788,
+      "step": 2240
+    },
+    {
+      "epoch": 2.4483133841131663,
+      "grad_norm": 0.03673616796731949,
+      "learning_rate": 1.750320062847116e-05,
+      "loss": 0.058925163745880124,
+      "step": 2250
+    },
+    {
+      "epoch": 2.4591947769314473,
+      "grad_norm": 0.050142526626586914,
+      "learning_rate": 1.6840723994175356e-05,
+      "loss": 0.05712746381759644,
+      "step": 2260
+    },
+    {
+      "epoch": 2.4700761697497278,
+      "grad_norm": 0.03750143572688103,
+      "learning_rate": 1.6189873144597267e-05,
+      "loss": 0.056266939640045165,
+      "step": 2270
+    },
+    {
+      "epoch": 2.4809575625680087,
+      "grad_norm": 0.03403429314494133,
+      "learning_rate": 1.5550739069608676e-05,
+      "loss": 0.05650065541267395,
+      "step": 2280
+    },
+    {
+      "epoch": 2.4918389553862896,
+      "grad_norm": 0.030664529651403427,
+      "learning_rate": 1.492341112105966e-05,
+      "loss": 0.05500165820121765,
+      "step": 2290
+    },
+    {
+      "epoch": 2.50272034820457,
+      "grad_norm": 0.03341777250170708,
+      "learning_rate": 1.4307977000287064e-05,
+      "loss": 0.05969707369804382,
+      "step": 2300
+    },
+    {
+      "epoch": 2.513601741022851,
+      "grad_norm": 0.03309670463204384,
+      "learning_rate": 1.3704522745853743e-05,
+      "loss": 0.05166618824005127,
+      "step": 2310
+    },
+    {
+      "epoch": 2.5244831338411315,
+      "grad_norm": 0.03246740996837616,
+      "learning_rate": 1.311313272152025e-05,
+      "loss": 0.0567930817604065,
+      "step": 2320
+    },
+    {
+      "epoch": 2.5353645266594125,
+      "grad_norm": 0.036715373396873474,
+      "learning_rate": 1.2533889604450677e-05,
+      "loss": 0.05777972936630249,
+      "step": 2330
+    },
+    {
+      "epoch": 2.546245919477693,
+      "grad_norm": 0.03496614098548889,
+      "learning_rate": 1.196687437365427e-05,
+      "loss": 0.054594647884368894,
+      "step": 2340
+    },
+    {
+      "epoch": 2.557127312295974,
+      "grad_norm": 0.03273237124085426,
+      "learning_rate": 1.141216629866434e-05,
+      "loss": 0.05491857528686524,
+      "step": 2350
+    },
+    {
+      "epoch": 2.568008705114255,
+      "grad_norm": 0.03323022276163101,
+      "learning_rate": 1.0869842928456441e-05,
+      "loss": 0.05289023518562317,
+      "step": 2360
+    },
+    {
+      "epoch": 2.5788900979325353,
+      "grad_norm": 0.042828429490327835,
+      "learning_rate": 1.033998008060687e-05,
+      "loss": 0.05806971788406372,
+      "step": 2370
+    },
+    {
+      "epoch": 2.5897714907508163,
+      "grad_norm": 0.03340109810233116,
+      "learning_rate": 9.822651830693152e-06,
+      "loss": 0.05867668390274048,
+      "step": 2380
+    },
+    {
+      "epoch": 2.6006528835690967,
+      "grad_norm": 0.03182734549045563,
+      "learning_rate": 9.317930501938289e-06,
+      "loss": 0.05183411836624145,
+      "step": 2390
+    },
+    {
+      "epoch": 2.6115342763873777,
+      "grad_norm": 0.03279723972082138,
+      "learning_rate": 8.82588665509978e-06,
+      "loss": 0.06119837760925293,
+      "step": 2400
+    },
+    {
+      "epoch": 2.6115342763873777,
+      "eval_loss": 0.06727322936058044,
+      "eval_runtime": 5.4288,
+      "eval_samples_per_second": 14.736,
+      "eval_steps_per_second": 1.842,
+      "step": 2400
+    },
+    {
+      "epoch": 2.622415669205658,
+      "grad_norm": 0.03709235042333603,
+      "learning_rate": 8.346589078605215e-06,
+      "loss": 0.05637891292572021,
+      "step": 2410
+    },
+    {
+      "epoch": 2.633297062023939,
+      "grad_norm": 0.030506664887070656,
+      "learning_rate": 7.880104778935505e-06,
+      "loss": 0.05910442471504211,
+      "step": 2420
+    },
+    {
+      "epoch": 2.64417845484222,
+      "grad_norm": 0.03575789928436279,
+      "learning_rate": 7.426498971257278e-06,
+      "loss": 0.054590433835983276,
+      "step": 2430
+    },
+    {
+      "epoch": 2.6550598476605005,
+      "grad_norm": 0.035656314343214035,
+      "learning_rate": 6.985835070305768e-06,
+      "loss": 0.056293058395385745,
+      "step": 2440
+    },
+    {
+      "epoch": 2.6659412404787814,
+      "grad_norm": 0.03722505271434784,
+      "learning_rate": 6.5581746815192135e-06,
+      "loss": 0.058118653297424314,
+      "step": 2450
+    },
+    {
+      "epoch": 2.676822633297062,
+      "grad_norm": 0.035609811544418335,
+      "learning_rate": 6.143577592426453e-06,
+      "loss": 0.05446176528930664,
+      "step": 2460
+    },
+    {
+      "epoch": 2.687704026115343,
+      "grad_norm": 0.035058531910181046,
+      "learning_rate": 5.742101764288521e-06,
+      "loss": 0.05432036519050598,
+      "step": 2470
+    },
+    {
+      "epoch": 2.6985854189336234,
+      "grad_norm": 0.036867234855890274,
+      "learning_rate": 5.353803323995521e-06,
+      "loss": 0.05071129202842713,
+      "step": 2480
+    },
+    {
+      "epoch": 2.7094668117519043,
+      "grad_norm": 0.030480151996016502,
+      "learning_rate": 4.9787365562200985e-06,
+      "loss": 0.04942301511764526,
+      "step": 2490
+    },
+    {
+      "epoch": 2.7203482045701852,
+      "grad_norm": 0.0435580313205719,
+      "learning_rate": 4.616953895828269e-06,
+      "loss": 0.05862914323806763,
+      "step": 2500
+    },
+    {
+      "epoch": 2.7312295973884657,
+      "grad_norm": 0.03596170246601105,
+      "learning_rate": 4.268505920549016e-06,
+      "loss": 0.054695284366607665,
+      "step": 2510
+    },
+    {
+      "epoch": 2.742110990206746,
+      "grad_norm": 0.034583065658807755,
+      "learning_rate": 3.933441343903444e-06,
+      "loss": 0.04879617691040039,
+      "step": 2520
+    },
+    {
+      "epoch": 2.752992383025027,
+      "grad_norm": 0.03540559113025665,
+      "learning_rate": 3.61180700839453e-06,
+      "loss": 0.05482098460197449,
+      "step": 2530
+    },
+    {
+      "epoch": 2.763873775843308,
+      "grad_norm": 0.0387350358068943,
+      "learning_rate": 3.3036478789585044e-06,
+      "loss": 0.05537814497947693,
+      "step": 2540
+    },
+    {
+      "epoch": 2.7747551686615886,
+      "grad_norm": 0.03608511760830879,
+      "learning_rate": 3.0090070366786792e-06,
+      "loss": 0.05486249327659607,
+      "step": 2550
+    },
+    {
+      "epoch": 2.7856365614798695,
+      "grad_norm": 0.03293207287788391,
+      "learning_rate": 2.7279256727626546e-06,
+      "loss": 0.054778552055358885,
+      "step": 2560
+    },
+    {
+      "epoch": 2.7965179542981504,
+      "grad_norm": 0.03599146753549576,
+      "learning_rate": 2.4604430827837143e-06,
+      "loss": 0.05409228205680847,
+      "step": 2570
+    },
+    {
+      "epoch": 2.807399347116431,
+      "grad_norm": 0.03957504406571388,
+      "learning_rate": 2.2065966611872657e-06,
+      "loss": 0.06026080846786499,
+      "step": 2580
+    },
+    {
+      "epoch": 2.8182807399347114,
+      "grad_norm": 0.033959709107875824,
+      "learning_rate": 1.9664218960630533e-06,
+      "loss": 0.05407561063766479,
+      "step": 2590
+    },
+    {
+      "epoch": 2.8291621327529923,
+      "grad_norm": 0.04043720290064812,
+      "learning_rate": 1.7399523641838278e-06,
+      "loss": 0.0522362232208252,
+      "step": 2600
+    },
+    {
+      "epoch": 2.8291621327529923,
+      "eval_loss": 0.06709529459476471,
+      "eval_runtime": 5.6005,
+      "eval_samples_per_second": 14.284,
+      "eval_steps_per_second": 1.786,
+      "step": 2600
+    },
+    {
+      "epoch": 2.8400435255712733,
+      "grad_norm": 0.034862909466028214,
+      "learning_rate": 1.5272197263113131e-06,
+      "loss": 0.050741416215896604,
+      "step": 2610
+    },
+    {
+      "epoch": 2.8509249183895538,
+      "grad_norm": 0.03554687649011612,
+      "learning_rate": 1.3282537227699787e-06,
+      "loss": 0.054420065879821775,
+      "step": 2620
+    },
+    {
+      "epoch": 2.8618063112078347,
+      "grad_norm": 0.0366034209728241,
+      "learning_rate": 1.143082169289289e-06,
+      "loss": 0.05170666575431824,
+      "step": 2630
+    },
+    {
+      "epoch": 2.8726877040261156,
+      "grad_norm": 0.03735072910785675,
+      "learning_rate": 9.717309531150132e-07,
+      "loss": 0.05548094511032105,
+      "step": 2640
+    },
+    {
+      "epoch": 2.883569096844396,
+      "grad_norm": 0.034383635967969894,
+      "learning_rate": 8.142240293901649e-07,
+      "loss": 0.05571337938308716,
+      "step": 2650
+    },
+    {
+      "epoch": 2.8944504896626766,
+      "grad_norm": 0.03764410689473152,
+      "learning_rate": 6.705834178060589e-07,
+      "loss": 0.05904418230056763,
+      "step": 2660
+    },
+    {
+      "epoch": 2.9053318824809575,
+      "grad_norm": 0.03647703677415848,
+      "learning_rate": 5.408291995238956e-07,
+      "loss": 0.0517579436302185,
+      "step": 2670
+    },
+    {
+      "epoch": 2.9162132752992385,
+      "grad_norm": 0.033308230340480804,
+      "learning_rate": 4.249795143673962e-07,
+      "loss": 0.05420621633529663,
+      "step": 2680
+    },
+    {
+      "epoch": 2.927094668117519,
+      "grad_norm": 0.03559701517224312,
+      "learning_rate": 3.230505582868637e-07,
+      "loss": 0.060290944576263425,
+      "step": 2690
+    },
+    {
+      "epoch": 2.9379760609358,
+      "grad_norm": 0.03408787399530411,
+      "learning_rate": 2.3505658109487238e-07,
+      "loss": 0.05462511777877808,
+      "step": 2700
+    },
+    {
+      "epoch": 2.9488574537540804,
+      "grad_norm": 0.036027781665325165,
+      "learning_rate": 1.6100988447423915e-07,
+      "loss": 0.06044437885284424,
+      "step": 2710
+    },
+    {
+      "epoch": 2.9597388465723613,
+      "grad_norm": 0.03578963875770569,
+      "learning_rate": 1.0092082025815464e-07,
+      "loss": 0.0504527747631073,
+      "step": 2720
+    },
+    {
+      "epoch": 2.970620239390642,
+      "grad_norm": 0.03743607550859451,
+      "learning_rate": 5.479778898302979e-08,
+      "loss": 0.053725385665893556,
+      "step": 2730
+    },
+    {
+      "epoch": 2.9815016322089227,
+      "grad_norm": 0.03584102541208267,
+      "learning_rate": 2.2647238714024188e-08,
+      "loss": 0.05662902593612671,
+      "step": 2740
+    },
+    {
+      "epoch": 2.9923830250272037,
+      "grad_norm": 0.029956763610243797,
+      "learning_rate": 4.473664143667122e-09,
+      "loss": 0.0520307719707489,
+      "step": 2750
+    },
+    {
+      "epoch": 3.0,
+      "step": 2757,
+      "total_flos": 4.480981168874848e+18,
+      "train_loss": 0.07706532943797881,
+      "train_runtime": 10631.5322,
+      "train_samples_per_second": 4.149,
+      "train_steps_per_second": 0.259
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 2757,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 200,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.480981168874848e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}

qwen3_8b_sft_rtl/training_args.bin

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:63e1337140001fd786f5bfde65ea532ba76972f9e8e0826ca7b80b663dbc5302
+size 5176